Academia.eduAcademia.edu

Outline

Title
Abstract
Figures
Case Extraction
Conclusion and Future Work
Rough Set Theory Methods for Feature Selection
Advantages of Rough Set Theory Methods
Proposed Feature Selection Method
Feature Selection Methods Comparison
Materials
Methodology
Results
System Design
Case-Based Reasoning
Case Structure
Case Retrieval
Nearest Neighbor Selection Case Matching
Case Generation
Case
Case Studies
Discussion and Concluding Remarks
Case-Based Reasoning as an Explanation Mechanism
Summary and Further Work
Discussion of Related Work
Processed Data
Summary and Outlook
Populating the Case Base
Experimental Evaluation
Conclusions and Future Work
Enabling Case-Based Reasoning on the Web of Data
Related Work
Background: Linked Data and the Web of Data
Case-Based View on the Web of Data
¦ ¥ 7 Discussion
On Retaining Web Search Cases
Web Search Case Retention
Reusing Captured Cases
Case-Based, Affordance-Guided Web
Population of the Case Base
Algorithm 1 Population of Case Base
Procedure
Queries / New Cases
Evaluation Methodology
Recommendations
Discussion
Acknowledgement
Workshop Overview
Results and Examples
Conclusions
Recipe Adaptation Using Formal Concept Analysis
Adaptation Cases for Cooking Workflows
Applying Adaptation Cases to New Situations
Experimental Evaluation and Discussion
Experimental Results
Conclusion
Introduction
References

Conversational framework for web search and recommendations

Profile image of Ashwin RamAshwin Ram

2010, ICCBR Workshop Proceedings

Abstract

Abstract. In this paper, we describe a Conversational Interaction framework as an innovative and natural approach to facilitate easier information access by combining web search and recommendations. This framework includes an intelligent information agent (Cobot) in the conversation that provides contextually relevant social and web search recommendations. This setup leverages the information discovery process by integrating web information retrieval along with proactive connections to relevant users who can participate in real ...

Figures (75)
Welcoming you, from left to right, are Isabelle Bichindaritz, Cindy Marling, and Stefania Montani. Photo by Peter Funk
Welcoming you, from left to right, are Isabelle Bichindaritz, Cindy Marling, and Stefania Montani. Photo by Peter Funk
Fig. 1. The proposed behavior modification architecture.
Fig. 1. The proposed behavior modification architecture.
Fig. 2. The figure shows the FSM corresponding to the failure pattern VCRGLfail. STRT, END represent the status of the behavior. PARAMInappr. represents the rou- tine that checks whether the location where peasant is gathering resources is at a location that is too close to the enemy compared to other resource gathering locations.
Fig. 2. The figure shows the FSM corresponding to the failure pattern VCRGLfail. STRT, END represent the status of the behavior. PARAMInappr. represents the rou- tine that checks whether the location where peasant is gathering resources is at a location that is too close to the enemy compared to other resource gathering locations.
Table 1. Some example failure patterns and their associated behavior modificatior operators in WARGUS
Table 1. Some example failure patterns and their associated behavior modificatior operators in WARGUS
Fig. 3. The figure shows the average results from experiments turning the behavior adaptation on and off
Fig. 3. The figure shows the average results from experiments turning the behavior adaptation on and off
Fig. 4. The figure shows the overall percentage improvement using the behavior adap- tation approach
Fig. 4. The figure shows the overall percentage improvement using the behavior adap- tation approach
Fig. 1. A case in D2 consisting of a plan, goal and game state. The snippet contains two actions and the game state representation is not fully included due to space limitations.
Fig. 1. A case in D2 consisting of a plan, goal and game state. The snippet contains two actions and the game state representation is not fully included due to space limitations.
Fig. 2. Our goal ontology for actions in StarCraft
Fig. 2. Our goal ontology for actions in StarCraft
Inputs: observations, conditionFeatures, decisionFeatures, maxPR Outputs: coreFeatures As for CBR performance, we do not expect statistically significant differences, or important differences otherwise. The same hypothesis holds for the number of selected features.
Inputs: observations, conditionFeatures, decisionFeatures, maxPR Outputs: coreFeatures As for CBR performance, we do not expect statistically significant differences, or important differences otherwise. The same hypothesis holds for the number of selected features.
Feature Selection Performance: Execution times are in minutes. The next two tables show the averaged metrics for CBR performance and Feature Selection performance, as well as the result of the ANOVA performed for the effect of the feature selection algorithm on the collected metrics.
Feature Selection Performance: Execution times are in minutes. The next two tables show the averaged metrics for CBR performance and Feature Selection performance, as well as the result of the ANOVA performed for the effect of the feature selection algorithm on the collected metrics.
Figure 1. System Architecture and Overview
Figure 1. System Architecture and Overview
Qur micro-level decision CBR project implementation is an extension of the work performed by Aamodt and Szczepanski's micro-level CBR system but is unique in that it integrates Reinforcement Learning (RL) techniques and a fully automated case generation framework meaning case adaptation and creation can occur online during a live game with an opponent. RL is a form of machine learning that can be used in games to allow agents to learn a policy/action to select during every state of the game. Our use of spatial temporal reasoning is fundamentally different from Aha et al.’s im- plementation (much like Sharma et al.) as our revision steps do not require pre- computed state lattices [1]. Sharma et al.’s implementation is also largely different from ours as they focus on the transference of generalized case learning between mul- tiple layers of a hierarchical implementation which offers delayed online results [7]. The goal of RL in our system is to arrive at an optimal action for each state by max- imizing the reward received through a process of both real-time and intermittent trial and error exploration feedback [5]. Using CBR coupled with RL eliminates the need for human feedback loop to learn effective strategies in the game. Additionally, our project integrates concepts from the fully integrated agent manager framework of Ma- teas and Weber [6]. However, our unit managers are implemented on a smaller scale due to the requirements of small scale combat. Instead of utilizing a large network of managers, our implementation only requires the use of a scout and group manager to perform the basic micro-level tactical behaviors of individual units [6]. 3 System Design
Qur micro-level decision CBR project implementation is an extension of the work performed by Aamodt and Szczepanski's micro-level CBR system but is unique in that it integrates Reinforcement Learning (RL) techniques and a fully automated case generation framework meaning case adaptation and creation can occur online during a live game with an opponent. RL is a form of machine learning that can be used in games to allow agents to learn a policy/action to select during every state of the game. Our use of spatial temporal reasoning is fundamentally different from Aha et al.’s im- plementation (much like Sharma et al.) as our revision steps do not require pre- computed state lattices [1]. Sharma et al.’s implementation is also largely different from ours as they focus on the transference of generalized case learning between mul- tiple layers of a hierarchical implementation which offers delayed online results [7]. The goal of RL in our system is to arrive at an optimal action for each state by max- imizing the reward received through a process of both real-time and intermittent trial and error exploration feedback [5]. Using CBR coupled with RL eliminates the need for human feedback loop to learn effective strategies in the game. Additionally, our project integrates concepts from the fully integrated agent manager framework of Ma- teas and Weber [6]. However, our unit managers are implemented on a smaller scale due to the requirements of small scale combat. Instead of utilizing a large network of managers, our implementation only requires the use of a scout and group manager to perform the basic micro-level tactical behaviors of individual units [6]. 3 System Design
Figure 3. Average Number of Cases Generated per Rush Test Set and only attacking when the opponent comes in range. As the case generation thre- shold increases (meaning cases become more generic) we can see that the win percen- tage increases against all strategies with exception of the random strategy opponent Our highest win percentage comes from the games against the AI opponent that uti- lized the rush strategy, however, our implementation was only able to win 45% of the games. The crucial piece of information to take away from this table is that as the case generation thresholds increase our AI walked away with more wins (we will elaborate on possible correlations in the following sections). After playing against the opposi- tion for roughly 8-12 games our AI would learn that it was best to draw a game end- ing in a stalemate rather than to keep assaulting the heavily defended positions of the enemy. Additionally, as the case generation thresholds reach 3000 our AI becomes aware of this fact much sooner than with lower thresholds.
Figure 3. Average Number of Cases Generated per Rush Test Set and only attacking when the opponent comes in range. As the case generation thre- shold increases (meaning cases become more generic) we can see that the win percen- tage increases against all strategies with exception of the random strategy opponent Our highest win percentage comes from the games against the AI opponent that uti- lized the rush strategy, however, our implementation was only able to win 45% of the games. The crucial piece of information to take away from this table is that as the case generation thresholds increase our AI walked away with more wins (we will elaborate on possible correlations in the following sections). After playing against the opposi- tion for roughly 8-12 games our AI would learn that it was best to draw a game end- ing in a stalemate rather than to keep assaulting the heavily defended positions of the enemy. Additionally, as the case generation thresholds reach 3000 our AI becomes aware of this fact much sooner than with lower thresholds.
In looking at the number of cases that were generated during each test run we can see in table 2 the number of cases created decreases as the case generation threshold increases. This is to be expected because as cases become more generic they are more likely to be reused over a larger set of problem situation instances. In figure 3 we can see the total number of cases generated and we observe this case generation process over a period of time. Looking at the cases generated in the 1000 threshold we notice that cases never stop being generated nor do they show any sign
In looking at the number of cases that were generated during each test run we can see in table 2 the number of cases created decreases as the case generation threshold increases. This is to be expected because as cases become more generic they are more likely to be reused over a larger set of problem situation instances. In figure 3 we can see the total number of cases generated and we observe this case generation process over a period of time. Looking at the cases generated in the 1000 threshold we notice that cases never stop being generated nor do they show any sign
Figure 5. Average Efficiency vs. Learning Comparison in Rush test sets eae aaa, > a aaa <> elena If we now focus on the efficiency values for each case generation threshold in our best performing test runs against the rush opponent strategy we find that in fig. 4 the case 1000 generation threshold level produces an always decreasing efficiency trend line which further makes it less viable for a challenging game opponent. Looking at the 2000 threshold level we notice that the line is always increasing suggesting that further tests should be performed on this threshold level to see the inevitable peak ef- ficiency value. At the 3000 threshold it becomes clear that our AI was actually over trained and that further learning should have been withheld somewhere around the 10th or 11th game as its efficiency peaked very quickly.
Figure 5. Average Efficiency vs. Learning Comparison in Rush test sets eae aaa, > a aaa <> elena If we now focus on the efficiency values for each case generation threshold in our best performing test runs against the rush opponent strategy we find that in fig. 4 the case 1000 generation threshold level produces an always decreasing efficiency trend line which further makes it less viable for a challenging game opponent. Looking at the 2000 threshold level we notice that the line is always increasing suggesting that further tests should be performed on this threshold level to see the inevitable peak ef- ficiency value. At the 3000 threshold it becomes clear that our AI was actually over trained and that further learning should have been withheld somewhere around the 10th or 11th game as its efficiency peaked very quickly.
Fig. 1: General Make Me Play Me architecture
Fig. 1: General Make Me Play Me architecture
Fig. 1. Passive observation of an expert interacting with the environment Using such an approach, cases can be generated in an automated manner through passive observation. Using a passive approach, the expert, as shown in Figure 1, interacts with the environment and may not be aware it is being observed. The observer is able to view, and record, both the state of the envi- ronment and the actions performed by the expert so that they can later be used to construct cases.
Fig. 1. Passive observation of an expert interacting with the environment Using such an approach, cases can be generated in an automated manner through passive observation. Using a passive approach, the expert, as shown in Figure 1, interacts with the environment and may not be aware it is being observed. The observer is able to view, and record, both the state of the envi- ronment and the actions performed by the expert so that they can later be used to construct cases.
Fig. 1. Trace-Based System: a general architecture. User produces traces stored in the TBMS by interracting with an observed application. The trace-based application uses traces to solve problems and may ask the TBMS to apply transformations. The solution is proposed to the user and this process is also recorded on the trace.
Fig. 1. Trace-Based System: a general architecture. User produces traces stored in the TBMS by interracting with an observed application. The trace-based application uses traces to solve problems and may ask the TBMS to apply transformations. The solution is proposed to the user and this process is also recorded on the trace.
puter recording of these activities. From a computer science point of view, this project aims at discovering knowledge from traces of activity. The ABSTRACT application starts by observing the activity by various means (i.e. various kinds of sensors) and collects a primary trace. The set of data collected is also at- tached to a moment of the activity identified by a “time-code”. The trace can then be manipulated and enriched with more abstract symbols through several transformations. Figure 2 and figure 3 illustrate the way traces are manipulated in ABSTRACT. Figure 2 shows three abstraction levels of the primary trace from a generic point of view. Figure 3 illustrates the same principle, but applied to a specific application domain: driver behavior analysis. Fig. 3. Abstract: an example of the user interface. The primary trace has been trans- formed to produce traces ” understandable” by users (e.g. analysts).
puter recording of these activities. From a computer science point of view, this project aims at discovering knowledge from traces of activity. The ABSTRACT application starts by observing the activity by various means (i.e. various kinds of sensors) and collects a primary trace. The set of data collected is also at- tached to a moment of the activity identified by a “time-code”. The trace can then be manipulated and enriched with more abstract symbols through several transformations. Figure 2 and figure 3 illustrate the way traces are manipulated in ABSTRACT. Figure 2 shows three abstraction levels of the primary trace from a generic point of view. Figure 3 illustrates the same principle, but applied to a specific application domain: driver behavior analysis. Fig. 3. Abstract: an example of the user interface. The primary trace has been trans- formed to produce traces ” understandable” by users (e.g. analysts).
Fig. 4. Copy/paste example. The trace of the mouse-based copy/paste (above) is com- pared to the trace of the keyboard-based copy/paste (below). At the highest abstraction level, both traces are identical. However they represent two different ways of performing the same task. Note that in both cases, the bottom traces are not primary traces, they are already transformed from other traces (for example ” Right Click + Copy Selection” has been transformed in ” Mouse-Based Copy”). level of abstraction is found, the second step consists of adapting the way of performing the action (” Copy”) by taking into account the user preferences (e.g. "only with the keyboard”). This example suggests an interesting observation: while most users use the mouse to perform a Copy/Paste, blind users may pre- fer less common solutions. Hence, provenance information may help a user to find the best available episodes for him, or at least to give him relevant adaptation knowledge (how to present “paste”: with mouse or with keyboard?). We think this issue should open new paths especially in adaptive systems and assistive technologies.
Fig. 4. Copy/paste example. The trace of the mouse-based copy/paste (above) is com- pared to the trace of the keyboard-based copy/paste (below). At the highest abstraction level, both traces are identical. However they represent two different ways of performing the same task. Note that in both cases, the bottom traces are not primary traces, they are already transformed from other traces (for example ” Right Click + Copy Selection” has been transformed in ” Mouse-Based Copy”). level of abstraction is found, the second step consists of adapting the way of performing the action (” Copy”) by taking into account the user preferences (e.g. "only with the keyboard”). This example suggests an interesting observation: while most users use the mouse to perform a Copy/Paste, blind users may pre- fer less common solutions. Hence, provenance information may help a user to find the best available episodes for him, or at least to give him relevant adaptation knowledge (how to present “paste”: with mouse or with keyboard?). We think this issue should open new paths especially in adaptive systems and assistive technologies.
Table 1. Condition and response Let us assume that the system now gets equipped with a barometer, which to some degree can be used to predict the weather. The new Bayesian network. including the barometer is depicted in Fig. 2. Using this new device the network is able to predict that cloudy weather in some situations will improve and become sunny. Thus, the system will sometimes suggest the user to go to the beach even when it is cloudy, for example when the barometer shows a trend towards higher
Table 1. Condition and response Let us assume that the system now gets equipped with a barometer, which to some degree can be used to predict the weather. The new Bayesian network. including the barometer is depicted in Fig. 2. Using this new device the network is able to predict that cloudy weather in some situations will improve and become sunny. Thus, the system will sometimes suggest the user to go to the beach even when it is cloudy, for example when the barometer shows a trend towards higher
Fig. 1. Bayesian network To exemplify the need for provenance of changes in a Bayesian network we wil use a very simple example. A simple Bayesian network, as depicted in Fig. 1, car be used to tell a user whether to go to the beach or not. Real world application would probably be much more complicated and even make recommendation without being explicitly asked to do so. However, for pedagogical reasons we will stick with simplicity.
Fig. 1. Bayesian network To exemplify the need for provenance of changes in a Bayesian network we wil use a very simple example. A simple Bayesian network, as depicted in Fig. 1, car be used to tell a user whether to go to the beach or not. Real world application would probably be much more complicated and even make recommendation without being explicitly asked to do so. However, for pedagogical reasons we will stick with simplicity.
Fig. 2. Modified Bayesian network pressure. This new observed behaviour is shown in the column marked ’New recommendation’ in Table 1.
Fig. 2. Modified Bayesian network pressure. This new observed behaviour is shown in the column marked ’New recommendation’ in Table 1.
Fig. 1. Example of provenance from the OPM specification, based on [18, p. 8 To illustrate OPM, Figure 1 shows an adaptation of a diagram of a prove- nance sample included in the OPM specification.! It represents a portion of the provenance trace for an image generated by a workflow used in the first prove- nance challenge, a competition testing the capabilities of systems to record and exploit provenance information [18]. Edges in this graph represent relationships between provenance entities, which are represented by various shapes. Ovals in this graph represent “artifacts,” specifically data, in this case. The rectangle rep- resents a “process” which both consumes and produces data. Finally, the pointed oval represents an “agent,” in this case a user who executed the workflow. Ad- ditionally, the process has been annotated with an execution time.
Fig. 1. Example of provenance from the OPM specification, based on [18, p. 8 To illustrate OPM, Figure 1 shows an adaptation of a diagram of a prove- nance sample included in the OPM specification.! It represents a portion of the provenance trace for an image generated by a workflow used in the first prove- nance challenge, a competition testing the capabilities of systems to record and exploit provenance information [18]. Edges in this graph represent relationships between provenance entities, which are represented by various shapes. Ovals in this graph represent “artifacts,” specifically data, in this case. The rectangle rep- resents a “process” which both consumes and produces data. Finally, the pointed oval represents an “agent,” in this case a user who executed the workflow. Ad- ditionally, the process has been annotated with an execution time.
Fig. 2. Example of internal provenance in CBR
Fig. 2. Example of internal provenance in CBR
Fig. 3. Solving the same target problem (P) in Lazier CBR after feedback on case C3 has been received. The shortest adaptation path from a confirmed case to the target problem is now C; >C,> P. Learning From Feedback. Fig. 3 illustrates the use of Lazier CBR to solve the same target problem (P) as in our previous example but in a slightly different context. The case base is the same as in Fig. 2 except that feedback on case C3 has been received from a reliable source, with the result that its status has changed from unconfirmed to confirmed. The shortest adaptation path from a confirmed case to the target problem, and therefore the one that will be used to solve the problem in Lazier CBR, is now C3 — C, > P. Note that the solution provided by Lazier CBR in this example takes ac- count of feedback on a case that would not be considered in a traditional CBR ap- proach to solving the target problem. It is also worth noting that problem-solving efficiency benefits from the feedback on case C3, not only in terms of search effort but also in terms of adaptation effort.
Fig. 3. Solving the same target problem (P) in Lazier CBR after feedback on case C3 has been received. The shortest adaptation path from a confirmed case to the target problem is now C; >C,> P. Learning From Feedback. Fig. 3 illustrates the use of Lazier CBR to solve the same target problem (P) as in our previous example but in a slightly different context. The case base is the same as in Fig. 2 except that feedback on case C3 has been received from a reliable source, with the result that its status has changed from unconfirmed to confirmed. The shortest adaptation path from a confirmed case to the target problem, and therefore the one that will be used to solve the problem in Lazier CBR, is now C3 — C, > P. Note that the solution provided by Lazier CBR in this example takes ac- count of feedback on a case that would not be considered in a traditional CBR ap- proach to solving the target problem. It is also worth noting that problem-solving efficiency benefits from the feedback on case C3, not only in terms of search effort but also in terms of adaptation effort.
Fig. 2. Solving a target problem (P) in Lazier CBR before any feedback is received. The short- est adaptation path from a confirmed case to the target problem is Cg > Cg > Cy > P.
Fig. 2. Solving a target problem (P) in Lazier CBR before any feedback is received. The short- est adaptation path from a confirmed case to the target problem is Cg > Cg > Cy > P.
Fig. 1. Knowledge Extraction Process Model for CBR Systems
Fig. 1. Knowledge Extraction Process Model for CBR Systems
Fig. 2. Screenshot of the Knowledge Extraction Workbench One the left hand side the knowledge engineer can browse forum posts and se- lect the analysis method. The right hand side displays the extracted taxonomies and allows the knowledge engineer to load and save different taxonomies as well as manipulate similarity measures. Further, the knowledge engineer can edit the taxonomy and recalculate the similarity assignments for any node within the taxonomy. A more detailed explanation of the workbench can be found in [18].
Fig. 2. Screenshot of the Knowledge Extraction Workbench One the left hand side the knowledge engineer can browse forum posts and se- lect the analysis method. The right hand side displays the extracted taxonomies and allows the knowledge engineer to load and save different taxonomies as well as manipulate similarity measures. Further, the knowledge engineer can edit the taxonomy and recalculate the similarity assignments for any node within the taxonomy. A more detailed explanation of the workbench can be found in [18].
Fig. 3. Taxonomy Depth Comparison
Fig. 3. Taxonomy Depth Comparison
Fig. 1. A screenshot from Ghost Writer-2.0
Fig. 1. A screenshot from Ghost Writer-2.0
Fig. 2. The number of suggestions that users directly incorporated into reviews
Fig. 2. The number of suggestions that users directly incorporated into reviews
Fig. 3. Noun phrases in reviews written with GW-2.0
Fig. 3. Noun phrases in reviews written with GW-2.0
Fig. 4. “The system helped me to write a comprehensive review”
Fig. 4. “The system helped me to write a comprehensive review”
Fig. 5. “I found the suggestions helpful”
Fig. 5. “I found the suggestions helpful”
Fig. 6. “The number of times I was inspired by suggestions but I didn’t click on them”
Fig. 6. “The number of times I was inspired by suggestions but I didn’t click on them”
Fig. 1. Overview of Linking Open Data sources as of July 2009 with source types shown
Fig. 1. Overview of Linking Open Data sources as of July 2009 with source types shown
Fig. 2. Example of a DBPedia resource with descriptions of the different vocabularies used for the properties
Fig. 2. Example of a DBPedia resource with descriptions of the different vocabularies used for the properties
Fig. 3. The CBR vocabulary and a case base with experiential data for user profiling. Properties from the CBR vocabulary are displayed with solid lines, all other properties are displayed in dashed lines.
Fig. 3. The CBR vocabulary and a case base with experiential data for user profiling. Properties from the CBR vocabulary are displayed with solid lines, all other properties are displayed in dashed lines.
Fig. 1. WebAdapt’s adaptation process and knowledge sources, from [8].
Fig. 1. WebAdapt’s adaptation process and knowledge sources, from [8].
Table 1. Query method, hierarchy nodes, and leaves for each source
Table 1. Query method, hierarchy nodes, and leaves for each source
Table 2. Performance improvements. All times measured in seconds. Fig. 2. Average mining time per problem for all configurations
Table 2. Performance improvements. All times measured in seconds. Fig. 2. Average mining time per problem for all configurations
Fig. 3. Case base growth Results for Question 2: Figure 3 shows the growth rates of the case base fot each method. When adding cases with unique features, the case base growth rate is nearly linear. Each knowledge source is capable of presenting dozens 01 hundreds of candidate substitutions for each adaptation, potentially leading tc extremely large case bases. The growth of the second method is nearly linear fo1 the first fifty-five adaptations, after which it starts to level off as constraints are re-encountered. The case base continues to grow slowly because some knowledge sources, such as OpenCyc and Wikipedia, are described by constraints which form a compact vocabulary. The constraint space method generated a total ot 397 cases compared to 625 for the candidate substitution method. Both methods had the same coverage.
Fig. 3. Case base growth Results for Question 2: Figure 3 shows the growth rates of the case base fot each method. When adding cases with unique features, the case base growth rate is nearly linear. Each knowledge source is capable of presenting dozens 01 hundreds of candidate substitutions for each adaptation, potentially leading tc extremely large case bases. The growth of the second method is nearly linear fo1 the first fifty-five adaptations, after which it starts to level off as constraints are re-encountered. The case base continues to grow slowly because some knowledge sources, such as OpenCyc and Wikipedia, are described by constraints which form a compact vocabulary. The constraint space method generated a total ot 397 cases compared to 625 for the candidate substitution method. Both methods had the same coverage.
Fig. 1. Blocking and affordance assignment approach in WebCBR The use of Case Based Reasoning (CBR) in web context has attracted researchers for a while. For example, Limthan et al. [6] applied CBR in order to compose a complex web service from heterogeneous web services residing in different parts of the web, and [2] in order to search and select web services. Ha [4] investigated how the web usage data can be used to discover navigation patterns which, in tum, can be used to predict the user behavior. In this work, a web document has a corresponding case representation capturing information that bridges queries to documents through affordances. Query experiences are used to adjust the affordances of documents. This provides search-useful and re-usable information for future web searches. The key rationale of the proposed approach is based on two assumptions: (i) a web document embraces a number of text blocks each of which can be connected to one or several affordances, (ii) the web can be searched CBR-wise and relevance of a document can be judged on the basis of its affordance alignment with the current query case. Figure. 1 illustrates the proposed approach in comparison with the web search.
Fig. 1. Blocking and affordance assignment approach in WebCBR The use of Case Based Reasoning (CBR) in web context has attracted researchers for a while. For example, Limthan et al. [6] applied CBR in order to compose a complex web service from heterogeneous web services residing in different parts of the web, and [2] in order to search and select web services. Ha [4] investigated how the web usage data can be used to discover navigation patterns which, in tum, can be used to predict the user behavior. In this work, a web document has a corresponding case representation capturing information that bridges queries to documents through affordances. Query experiences are used to adjust the affordances of documents. This provides search-useful and re-usable information for future web searches. The key rationale of the proposed approach is based on two assumptions: (i) a web document embraces a number of text blocks each of which can be connected to one or several affordances, (ii) the web can be searched CBR-wise and relevance of a document can be judged on the basis of its affordance alignment with the current query case. Figure. 1 illustrates the proposed approach in comparison with the web search.
Solution: The extracted text, say text;, of the block b; is processed to identify its affordance with the help of the resource terms (topics and # terms considered in each topic, are presented in the table 1). For each topic, the matching terms are identified, the affordance with respect to this topic is computed and the affordance vector of the block is updated (see ComputeBlockA ffordance in the Algorithm 2) accordingly. The AV of the document, in turn, is computed by ComputeDocA ffordance, as described in the ‘solution part’ of the Algorithm 1. A case is generated by coupling the problem description and the solution parts and is added to the case base.
Solution: The extracted text, say text;, of the block b; is processed to identify its affordance with the help of the resource terms (topics and # terms considered in each topic, are presented in the table 1). For each topic, the matching terms are identified, the affordance with respect to this topic is computed and the affordance vector of the block is updated (see ComputeBlockA ffordance in the Algorithm 2) accordingly. The AV of the document, in turn, is computed by ComputeDocA ffordance, as described in the ‘solution part’ of the Algorithm 1. A case is generated by coupling the problem description and the solution parts and is added to the case base.
unwanted, restricted contents, adult contents, etc), were selected for our experiments. We have manually crafted the list of 18 affordances related to tourism domain including the affordance miscellaneous.
unwanted, restricted contents, adult contents, etc), were selected for our experiments. We have manually crafted the list of 18 affordances related to tourism domain including the affordance miscellaneous.
Fig. 2. Effect of rank aggregation of lucene retrieval vs the proposed approach During the estimation of case affordance vectors, the values of the elements in the vector increase with the number of matching terms between the solution part and the new case. In such situations, we could apply affordance vector length normalization. To length normalize the elements of the affordance vector AV., for the case c having m affordances: {A;,,A;,,--- ,Az,, }, to the unit vector, we do the following: AV. := AV./|AV.| where denominator denotes the Euclidean length of the vector of the affordance ¢; in c. In the mean time, we will take care of the effect of normalization factor in decreasing the chances of retrieval of the document.
Fig. 2. Effect of rank aggregation of lucene retrieval vs the proposed approach During the estimation of case affordance vectors, the values of the elements in the vector increase with the number of matching terms between the solution part and the new case. In such situations, we could apply affordance vector length normalization. To length normalize the elements of the affordance vector AV., for the case c having m affordances: {A;,,A;,,--- ,Az,, }, to the unit vector, we do the following: AV. := AV./|AV.| where denominator denotes the Euclidean length of the vector of the affordance ¢; in c. In the mean time, we will take care of the effect of normalization factor in decreasing the chances of retrieval of the document.
Fig. 3. Effect of the proposed approach on a few selected queries
Fig. 3. Effect of the proposed approach on a few selected queries
Figure 1 gives a high level architecture of the Conversational Interaction framework. The framework is built around constructs to support memory update and access, cat- egorization and learning in the system. The framework allows for the ability to start conversations, get connected to people and get relevant information for the information need in context. While developing the Conversational Interaction framework, we are adhering to some requirements and guidelines which are as follows:
Figure 1 gives a high level architecture of the Conversational Interaction framework. The framework is built around constructs to support memory update and access, cat- egorization and learning in the system. The framework allows for the ability to start conversations, get connected to people and get relevant information for the information need in context. While developing the Conversational Interaction framework, we are adhering to some requirements and guidelines which are as follows:
Text Analysis and Processing Engine(TAPE) processes conversations, pushing i hrough the various steps of analysis, processing and storage within the system. The Surrent system is being designed and developed for health domain and engages in it the ise of medical ontologies coupled with natural language processing components. TAPE Figure 2) produces and maintains the knowledge representation by processing infor mation from agent’s working memory of conversation, user models and knowledge pases. The agent’s task is to use the sub-modules for extracting meaningful querie: from conversations, classifying messages into relevant categories, and calling the righ combination of algorithms for retrieving candidate recommendations.
Text Analysis and Processing Engine(TAPE) processes conversations, pushing i hrough the various steps of analysis, processing and storage within the system. The Surrent system is being designed and developed for health domain and engages in it the ise of medical ontologies coupled with natural language processing components. TAPE Figure 2) produces and maintains the knowledge representation by processing infor mation from agent’s working memory of conversation, user models and knowledge pases. The agent’s task is to use the sub-modules for extracting meaningful querie: from conversations, classifying messages into relevant categories, and calling the righ combination of algorithms for retrieving candidate recommendations.
The purpose of Episodic Memory is to capture the user’s short-term interactions and interests. Based on user’s frequency of interactions and diversity in topics, this memory empirically varies in the range of a few days for different users. The Semantic Memory captures the user’s long-term profile. These are the topics that interest the user in general and for a prolonged time. These interests change less frequently and represent general criteria of recommendation to the user. Many times, users might be interested in some temporary information need. Such information need not be incorporated in the long term user memory. The episodic memory captures such short-term interests. The episodic memory forms a sort of staging area and the concepts from this memory are selectively and periodically moved to the semantic memory in a crossover process. interchangeably use the terms Cobot and the Conversational Interaction framework) leverages these interactions to maintain users’ episodic and long term semantic models, agent’s per conversation working memory of topics, users and messages (Figure 3). The agent analyzes these memory structures to bring in external recommendations into the system by matching with the contextual information need(Categorization). The social feedback on the recommendations are registered in the indices for the algorithms to generate their contextual relevance. Paper [13] also describes the architecture of Cobot System in more detail.
The purpose of Episodic Memory is to capture the user’s short-term interactions and interests. Based on user’s frequency of interactions and diversity in topics, this memory empirically varies in the range of a few days for different users. The Semantic Memory captures the user’s long-term profile. These are the topics that interest the user in general and for a prolonged time. These interests change less frequently and represent general criteria of recommendation to the user. Many times, users might be interested in some temporary information need. Such information need not be incorporated in the long term user memory. The episodic memory captures such short-term interests. The episodic memory forms a sort of staging area and the concepts from this memory are selectively and periodically moved to the semantic memory in a crossover process. interchangeably use the terms Cobot and the Conversational Interaction framework) leverages these interactions to maintain users’ episodic and long term semantic models, agent’s per conversation working memory of topics, users and messages (Figure 3). The agent analyzes these memory structures to bring in external recommendations into the system by matching with the contextual information need(Categorization). The social feedback on the recommendations are registered in the indices for the algorithms to generate their contextual relevance. Paper [13] also describes the architecture of Cobot System in more detail.
Table 1. Message Classification
Table 1. Message Classification
There are many technical challenges in community based information and recom- mendation systems. Cobot is being developed around the principle of ‘Suit the user, make it easy, make it good’. Cobot’s approach and solution to next generation of socially enabled search is uniquely driven by new trends on the web, requiring new technologies for an integrated socio-semantic search experience. Instead of relying on search engines that inundate the user with a multitude of information, Cobot models the information finding task as an interactive and collaborative recommendation process within a social community. The user describes his need in natural language to a trusted community which is modeled via text conversations familiar to most users. Our agent based con- versational framework for web search and recommendations uses a “wisdom of crowds’ approach to compensate for the limitations of traditional search engines and uses the ex- perience of real users by proactively bringing them to participate in the conversations.
There are many technical challenges in community based information and recom- mendation systems. Cobot is being developed around the principle of ‘Suit the user, make it easy, make it good’. Cobot’s approach and solution to next generation of socially enabled search is uniquely driven by new trends on the web, requiring new technologies for an integrated socio-semantic search experience. Instead of relying on search engines that inundate the user with a multitude of information, Cobot models the information finding task as an interactive and collaborative recommendation process within a social community. The user describes his need in natural language to a trusted community which is modeled via text conversations familiar to most users. Our agent based con- versational framework for web search and recommendations uses a “wisdom of crowds’ approach to compensate for the limitations of traditional search engines and uses the ex- perience of real users by proactively bringing them to participate in the conversations.
Fig. 1. JADAWeb Web-Based Graphical User Interface The web interface allows the user writing a natural language sentence with the recipe requirements (see Figure 1). The following is an example of a query where the user explains what (s)he wants: ”I want to eat some fruits, like apple, but I don’t like kiwi”. The system connects to the dependency text parser to obtain the information contained in this query. It extracts the list of ingredients to include and to avoid in the recipe.
Fig. 1. JADAWeb Web-Based Graphical User Interface The web interface allows the user writing a natural language sentence with the recipe requirements (see Figure 1). The following is an example of a query where the user explains what (s)he wants: ”I want to eat some fruits, like apple, but I don’t like kiwi”. The system connects to the dependency text parser to obtain the information contained in this query. It extracts the list of ingredients to include and to avoid in the recipe.
Fig. 2. Dependency tree The evaluation results show that MINIPAR is able to cover about 79 % of the dependency relationships in the SUSANNE corpus[10] with about 89 % precision(8] for English parsing, so we have a small quantity of errors, but it does not mean that our algorithm committed 10% of errors because is possible that the errors make by Minipar are not important for our purposes.
Fig. 2. Dependency tree The evaluation results show that MINIPAR is able to cover about 79 % of the dependency relationships in the SUSANNE corpus[10] with about 89 % precision(8] for English parsing, so we have a small quantity of errors, but it does not mean that our algorithm committed 10% of errors because is possible that the errors make by Minipar are not important for our purposes.
Fig. 3. Fragment of the ontology using by JadaWeb, using two examples of branches within the tree.
Fig. 3. Fragment of the ontology using by JadaWeb, using two examples of branches within the tree.
Table 1. Some examples of the fuzzy table. Another novelty in JadaWeb is a table of similarity between fuzzy ingredients. It is a table with specific values (fuzzy) that gives meaning to say that rice is replaceable by macaroni with a value of 0.4 to add to the similarity function. All these pairs of ingredients are defined in an XML file which indicates what ingredients are similar and the associated fuzzy value. The fuzzy table is not only used to replace pairs of ingredients. The system also can decide replacing one ingredient by a group of similar ingredients: stock instead of water and bouillon cubes with a similarity of 0.9, milk instead of water and cream in a similarity value of 0.6.
Table 1. Some examples of the fuzzy table. Another novelty in JadaWeb is a table of similarity between fuzzy ingredients. It is a table with specific values (fuzzy) that gives meaning to say that rice is replaceable by macaroni with a value of 0.4 to add to the similarity function. All these pairs of ingredients are defined in an XML file which indicates what ingredients are similar and the associated fuzzy value. The fuzzy table is not only used to replace pairs of ingredients. The system also can decide replacing one ingredient by a group of similar ingredients: stock instead of water and bouillon cubes with a similarity of 0.9, milk instead of water and cream in a similarity value of 0.6.
Table 2. Results obtained from the measure of evaluation If the similarity level increases then the number of substitutions also in- creases. With a similarity level that allows only adaptation between ingredients that share the same branch (similarity level = 1), the satisfaction with the giver recipe is very high. However, the result is worse when we use higher levels. This is because people do not expect those adapted elements. In various dishes (e.g. replacing the lemon with bread), the result is quite concerned, but we could see that a higher-level change (similarity level > 2) makes the user obtain another recipe recommended that he or she could not expect it. With a deeperd ontology. higher level adjustments would score much more satisfactory results. The result of 70 % for level 2 is good, but the 40 % of level 3 is very low.
Table 2. Results obtained from the measure of evaluation If the similarity level increases then the number of substitutions also in- creases. With a similarity level that allows only adaptation between ingredients that share the same branch (similarity level = 1), the satisfaction with the giver recipe is very high. However, the result is worse when we use higher levels. This is because people do not expect those adapted elements. In various dishes (e.g. replacing the lemon with bread), the result is quite concerned, but we could see that a higher-level change (similarity level > 2) makes the user obtain another recipe recommended that he or she could not expect it. With a deeperd ontology. higher level adjustments would score much more satisfactory results. The result of 70 % for level 2 is good, but the 40 % of level 3 is very low.
Fig. 4. Comparation with test 2. Comparing JADAWeb with Cookiis As a measure of the behaviour of our system, we have compared JADAWeb results with last year winner: CookIIS[7]. We show the similarities and differences between the two systems by comparing the retrieved sets of recipes of both systems. In the first query there is no coincidence between the retrieved recipes. JADAWeb retrieves Paella in the first place that is a highlight answer because we used the paella ingredients to make this query. Cookiis obtains Pad Thai in the first place. In the other queries we have some coincident results between the first 5 suggested recipes. In query 2 the two systems retrieve two common recipes, although in a different order. Idiot Bread Pizza is the first option in both. Pizza Quiche is the second in JADAWeb, but the fourth in Cookiis. In query 3 we only have one coincident result: Creamy Waldorf Salad.
Fig. 4. Comparation with test 2. Comparing JADAWeb with Cookiis As a measure of the behaviour of our system, we have compared JADAWeb results with last year winner: CookIIS[7]. We show the similarities and differences between the two systems by comparing the retrieved sets of recipes of both systems. In the first query there is no coincidence between the retrieved recipes. JADAWeb retrieves Paella in the first place that is a highlight answer because we used the paella ingredients to make this query. Cookiis obtains Pad Thai in the first place. In the other queries we have some coincident results between the first 5 suggested recipes. In query 2 the two systems retrieve two common recipes, although in a different order. Idiot Bread Pizza is the first option in both. Pizza Quiche is the second in JADAWeb, but the fourth in Cookiis. In query 3 we only have one coincident result: Creamy Waldorf Salad.
Fig. 3. Excerpt of the formal description of the preparation for “Glutinous rice with mangoes”.
Fig. 3. Excerpt of the formal description of the preparation for “Glutinous rice with mangoes”.
Fig. 5. Concept lattice corresponding to the context of Table 1. Each node (formal concept) is a set of objects O (extension, noted “E”) and a set of attributes A (intension, noted “T’’) such that all objects in O have all attributes in A in common in exclusion to any other. They are ordered according to the inclusion of their extension. is selected since, while both fig_53 and fig_1163 appear directly above mango, re- placing mango with fig_1163 would require removing three actions and adding four, whereas replacing it with fig_53 would require only removing three actions.
Fig. 5. Concept lattice corresponding to the context of Table 1. Each node (formal concept) is a set of objects O (extension, noted “E”) and a set of attributes A (intension, noted “T’’) such that all objects in O have all attributes in A in common in exclusion to any other. They are ordered according to the inclusion of their extension. is selected since, while both fig_53 and fig_1163 appear directly above mango, re- placing mango with fig_1163 would require removing three actions and adding four, whereas replacing it with fig_53 would require only removing three actions.
Fig. 3: Sample query to adapt a recipe for a beef noodle casserole.
Fig. 3: Sample query to adapt a recipe for a beef noodle casserole.
Fig. 5: Adaptation result (actually generated solution, some locations for changes could not be determined). Fig. 4: Adaptation result (optimal solution, ‘constructed by hand).
Fig. 5: Adaptation result (actually generated solution, some locations for changes could not be determined). Fig. 4: Adaptation result (optimal solution, ‘constructed by hand).
Fig. 6: Results of applying adaptation cases to new recipes in Experiment 2. As the retrieval is not part of this evaluation the best-fitting adaptation case is chosen for each query by hand. The resulting adapted workflow is compared to a solution obtained by manually transferring the respective adaptation steps to the workflow (see Fig. 6). Altogether, half of the test cases were adapted totally correctly. In the remaining cases, at least the data objects could be added correctly, while the correctness of added data links is only approximately 50%. Also only about 60% of the possible delete operations of data links are applied. This is a clear indication that the currently used syntactic similarity measure is insufficient for our purposes. Further experiments with improved similarity measures using the CookingCake’s ingredient ontology [2] as well as a task ontology (for preparation steps) will be necessary. Furthermore, the data flow could be extended by further data objects than ingredients like aggregated ingredients or utensils. However, the first results are quite promising and confirm that our idea of using the workflow paradigm to automatically adapt cooking instructions is feasible in principle.
Fig. 6: Results of applying adaptation cases to new recipes in Experiment 2. As the retrieval is not part of this evaluation the best-fitting adaptation case is chosen for each query by hand. The resulting adapted workflow is compared to a solution obtained by manually transferring the respective adaptation steps to the workflow (see Fig. 6). Altogether, half of the test cases were adapted totally correctly. In the remaining cases, at least the data objects could be added correctly, while the correctness of added data links is only approximately 50%. Also only about 60% of the possible delete operations of data links are applied. This is a clear indication that the currently used syntactic similarity measure is insufficient for our purposes. Further experiments with improved similarity measures using the CookingCake’s ingredient ontology [2] as well as a task ontology (for preparation steps) will be necessary. Furthermore, the data flow could be extended by further data objects than ingredients like aggregated ingredients or utensils. However, the first results are quite promising and confirm that our idea of using the workflow paradigm to automatically adapt cooking instructions is feasible in principle.
Fig. 1. The architecture of e:IAS used for CookIIS
Fig. 1. The architecture of e:IAS used for CookIIS
Fig. 2. The index and search pipeline for CookIIS Indexing and Searching recipes in CookIIS is done using workflow-like constructs, called pipelines in e:IAS. The pipelines consist of the so-called pipelets which represent the different steps (actions) for each operation.
Fig. 2. The index and search pipeline for CookIIS Indexing and Searching recipes in CookIIS is done using workflow-like constructs, called pipelines in e:IAS. The pipelines consist of the so-called pipelets which represent the different steps (actions) for each operation.
Here the terms title and body are the key words indicating that the text strings of the recipe title and body follow as assigners. 3.7 Menu A term in Froglingo can be used to express a menu too. Here is a sample expression for a dinner: dinner (Potato Salad 2) (Spicy Parmesan Chicken) (Sorry Cake); However, enumerating all the possible dinner choices is not preferred. We choose to have variables for the set of dinners producible from dishes. The following sample expression is equivalent to a set of dinners including the one described earlier: cyt Fa a eo
Here the terms title and body are the key words indicating that the text strings of the recipe title and body follow as assigners. 3.7 Menu A term in Froglingo can be used to express a menu too. Here is a sample expression for a dinner: dinner (Potato Salad 2) (Spicy Parmesan Chicken) (Sorry Cake); However, enumerating all the possible dinner choices is not preferred. We choose to have variables for the set of dinners producible from dishes. The following sample expression is equivalent to a set of dinners including the one described earlier: cyt Fa a eo
What we have done so far is eventually to support a user interface such that the customers who don’t know the system representation of recipes can easily express their requests, and the system can precisely process the requests. The resulting system allows users to enter a list of phrases or words delimited by a comma, and each phrase
What we have done so far is eventually to support a user interface such that the customers who don’t know the system representation of recipes can easily express their requests, and the system can precisely process the requests. The resulting system allows users to enter a list of phrases or words delimited by a comma, and each phrase

Related papers

Collaborative Information Access: A Conversational Search Approach
Saurav Sahay

Abstract. Knowledge and user generated content is proliferating on the web in scientific publications, information portals and online social media. This knowledge explosion has continued to outpace technological innovation in efficient information access technologies. In this paper, we describe the methods and technologies for 'Conversational Search'as an innovative solution to facilitate easier information access and reduce the information overload for users.

View PDFchevron_right
Socio-semantic conversational information access
Ashwin Ram

We develop an innovative approach to delivering relevant information using a combination of socio-semantic search and filtering approaches. The goal is to facilitate timely and relevant information access through the medium of conversations by mixing past community specific conversational knowledge and web information access to recommend and connect users and information together. Conversational Information Access is a socio-semantic search and recommendation activity with the goal to interactively engage people in conversations by receiving agent supported recommendations. It is useful because people engage in online social discussions unlike solitary search; the agent brings in relevant information as well as identifies relevant users; participants provide feedback during the conversation that the agent uses to improve its recommendations.

View PDFchevron_right
A Domain-independent Framework for building Conversational Recommender Systems
Giovanni Semeraro

2018

Conversational Recommender Systems (CoRSs) implement a paradigm where users can interact with the system for defining their preferences and discovering items that best fit their needs. A CoRS can be straightforwardly implemented as a chatbot. Chatbots are becoming more and more popular for several applications like customer care, health care, medical diagnoses. In the most complex form, the implementation of a chatbot is a challenging task since it requires knowledge about natural language processing, human-computer interaction, and so on. In this paper, we propose a general framework for making easy the generation of conversational recommender systems. The framework, based on a contentbased recommendation algorithm, is independent from the domain. Indeed, it allows to build a conversational recommender system with different interaction modes (natural language, buttons, hybrid) for any domain. The framework has been evaluated on two state-of-the-art datasets with the aim of identify...

View PDFchevron_right
Automatic Generation of Chatbots for Conversational Web Browsing
Marcos Baez

Conceptual Modeling, 2020

In this paper, we describe the foundations for generating a chatbot out of a website equipped with simple, bot-specific HTML annotations. The approach is part of what we call conversational web browsing, i.e., a dialog-based, natural language interaction with websites. The goal is to enable users to use content and functionality accessible through rendered UIs by "talking to websites" instead of by operating the graphical UI using keyboard and mouse. The chatbot mediates between the user and the website, operates its graphical UI on behalf of the user, and informs the user about the state of interaction. We describe the conceptual vocabulary and annotation format, the supporting conversational middleware and techniques, and the implementation of a demo able to deliver conversational web browsing experiences through Amazon Alexa.

View PDFchevron_right
CLASSY: A Conversational Aware Suggestion System
Diogo Gomes

Proceedings, 2019

Over the last few years, pervasive systems have seen some interesting development. Nevertheless, human–human interaction can also take advantage of those systems by using their ability to perceive the surrounding environment. In this work, we have developed a pervasive system – named CLASSY – that is aware of the conversational context and suggests documents potentially useful to the users based on an Information Retrieval system, and proposed a new scoring approach that uses semantics and distance based on proximity data in order to classify the relationship between tokens.

View PDFchevron_right
Conversational Semantic Search
Alex Marin

Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining, 2018

User expectations of web search are changing. They are expecting search engines to answer questions, to be more conversational, and to offer means to complete tasks on their behalf. At the same time, to increase the breadth of tasks that personal digital assistants (PDAs), such as Microsoft's Cortana or Amazon's Alexa, are capable of, PDAs need to better utilize information about the world, a significant amount of which is available in the knowledge bases and answers built for search engines. It thus seems likely that the underlying systems that power web search and PDAs will converge. This demonstration presents a system that merges elements of traditional multi-turn dialog systems with web based question answering. This demo focuses on the automatic composition of semantic functional units, Botlets, to generate responses to user's natural language (NL) queries. We show that such a system can be trained to combine information from search engine answers with PDA tasks to enable new user experiences.

View PDFchevron_right
Exploring Current User Web Search Behaviours in Analysis Tasks to be Supported in Conversational Search
Abhishek Kaushik

2021

Conversational search presents opportunities to support users in their search activities to improve the effectiveness and efficiency of search while reducing their cognitive load. Limitations of the potential competency of conversational agents restrict the situations for which conversational search agents can replace human intermediaries. It is thus more interesting, initially at least, to investigate opportunities for conversational interaction to support less complex information retrieval tasks, such as typical web search, which do not require human level intelligence in the conversational agent. In order tomove towards the development of a system to enable conversational search of this type, we need to understand their required capabilities. To progress our understanding of these, we report a study examining the behaviour of users when using a standard web search engine, designed to enable us to identify opportunities to support their search activities using a conversational agent.

View PDFchevron_right
Recommendation as Collaboration in Web Search
Maurice Coyle

AI Magazine, 2011

Recommender systems now play an important role in online information discovery, complementing traditional approaches such as search and navigation, with a more proactive approach to discovery that is informed by the users interests and preferences. To date recommender systems have been deployed within a variety of e-commerce domains, covering a range of products such as books, music, movies, and have proven to be a successful way to convert browsers into buyers. Recommendation technologies have a potentially much greater role to play in information discovery however and in this article we consider recent research that takes a fresh look at web search as a fertile platform for recommender systems research as users demand a new generation of search engines that are less susceptible to manipulation and more responsive to searcher needs and preferences.

View PDFchevron_right
CHARM: A Group Recommender ChatBot
Hanif Emamgholizadeh

Adjunct Proceedings of the 31st ACM Conference on User Modeling, Adaptation and Personalization

Group recommender systems (GRSs) are tools that support a group to find items that the whole group would enjoy experiencing jointly. There are two main lines of research in this field. The first line of research focuses on methods that combine the preferences of individual group members to obtain a group preference model and generate appropriate recommendations. The second line of research is more holistic and aims to support groups in all the phases of their decision-making process. The majority of the approaches of the second type use a simple conversational approach, which is critiquing. However, nowadays people heavily rely on social and chat platforms to make group decisions, and we believe that these platforms could be a valuable mean for building more effective GRSs. To this end, we have designed a framework tool that extends standard chat platforms by augmenting it with a chat-bot. The chat-bot enables the communication between the users on one side and the group recommender agent on the other. Our goal is a new holistic approach to group recommendations that would be the more beneficial than previous proposed conversational approaches. We aim to provide the proposed framework as an open environment for researchers to prototype their own GRSs. CCS CONCEPTS • Information systems → Personalization; Decision support systems; • Human-centered computing → Interaction design process and methods; Interactive systems and tools.

View PDFchevron_right
A conversational agent for information retrieval based on a study of human dialogues
Alexandre Pauchet

Alexandre Pauchet. A conversational agent for information retrieval based on a study of human dialogues.

View PDFchevron_right

References (261)

  1. David Aha, Matthew Molineaux, and Marc Ponsen. Learning to win: Case-based plan selection in a real-time strategy game. In ICCBR'2005, number 3620 in Lecture Notes in Artificial Intelligence, pages 5-20, 2005.
  2. Tristan Cazenave. Metarules to improve tactical go knowledge. Inf. Sci. Inf. Comput. Sci., 154(3-4):173-188, 2003. References
  3. Buro, M.: Real-Time Strategy Games: A New AI Research Challenge. In: Pro- ceedings of the International Joint Conference on Artificial Intelligence. (2003) 1534-1535
  4. Aha, D.W., Molineaux, M., Ponsen, M.: Learning to Win: Case-Based Plan Se- lection in a Real-Time Strategy Game. Lecture notes in computer science 3620 (2005) 5-20
  5. Ontanón, S., Mishra, K., Sugandh, N., Ram, A.: On-Line Case-Based Planning. Computational Intelligence 26(1) (2010) 84-119
  6. Weber, B., Mateas, M.: Case-Based Reasoning for Build Order in Real-Time Strat- egy Games. In: Proceedings of the Artificial Intelligence and Interactive Digital Entertainment Conference, AAAI Press (2009) 106-111
  7. Ontañón, S., Bonnette, K., Mahindrakar, P., Gómez-Martín, M., Long, K., Rad- hakrishnan, J., Shah, R., Ram, A.: Learning from Human Demonstrations for Real-Time Case-Based Planning. IJCAI Workshop on Learning Structural Knowl- edge from Observations (2009)
  8. Orkin, J.: Applying Goal-Oriented Action Planning to Games. In: AI Game Programming Wisdom 2. Charles River Media, S. Rabin editor (2003) 217-228
  9. Dill, K., Papp, D.: A Goal-Based Architecture for Opposing Player AI. In: Proceed- ings of the Artificial Intelligence and Interactive Digital Entertainment Conference, AAAI Press (2005)
  10. Szczepański, T., Aamodt, A.: Case-based reasoning for improved micromanage- ment in real-time strategy games. In: Proceedings of the ICCBR 2009 Workshop on CBR for Computer Games. (2009)
  11. Weber, B., Mateas, M.: A Data Mining Approach to Strategy Prediction. In: Proceedings of the IEEE Symposium on Computational Intelligence and Games, IEEE Press (2009) 140-147
  12. Metoyer, R., Stumpf, S., Neumann, C., Dodge, J., Cao, J., Schnabel, A.: Explaining How to Play Real-Time Strategy Games. Research and Development in Intelligent Systems XXVI (2010) 249-262
  13. Team Liquid: Liquipedia: The StarCraft Encyclopedia (April 2010) http://wiki.teamliquid.net/starcraft.
  14. Murata, T.: Petri nets: Properties, Analysis and Applications. Proceedings of the IEEE 77(4) (1989) 541-580
  15. Hogg, C., Munoz-Avila, H., Kuter, U.: HTN-MAKER: Learning HTNs with Min- imal Additional Knowledge Engineering Required. In: Proceedings of the Twenty- Third AAAI Conference on Artificial Intelligence. (2008)
  16. Könik, T., Laird, J.E.: Learning goal hierarchies from structured observations and expert annotations. Mach. Learn. 64(1-3) (2006) 263-287
  17. Weber, B., Mateas, M., Jhala, A.: Case-Based Goal Formulation. In: Proceedings of the AAAI Workshop on Goal-Driven Autonomy. (2010)
  18. Ontañón, S., Mishra, K., Sugandh, N., Ram, A.: Learning from Demonstration and Case-Based Planning for Real-Time Strategy Games. Soft Computing Applications in Industry (2008) 293-310
  19. Liu, H., Motoda, H.: Less is more. In: Computational Methods of Feature Selection. Chapman & Hall/CRC, Boca Raton (2008)
  20. Pal, S.K., Shiu, S.C.K.: Introduction. In: Foundations of soft case-based reasoning. Wiley-Interscience, Hoboken, N.J. (2004)
  21. Floyd, M.W., Esfandiari, B., Lam, K.: A case-based reasoning approach to imi- tating RoboCup players. In: Proceedings of the Twenty-first International Florida Artificial Intelligence Research Society Conference, AAAI Press (2008) 251-256
  22. Wettschereck, D., Aha, D.: Weighting features. In: Proceedings of the First Inter- national Conference on Case-Based Reasoning, Springer (1995) 347-358
  23. Floyd, M.W.: Improving the performance of a RoboCup case-based imitation agent through preprocessing of the case base. Master's thesis, Carleton University, Ottawa, Ontario, Canada (December 2008)
  24. Zhong, N., Dong, J.Z., Ohsuga, S.: Using rough sets with heuristics for feature selection. Journal of Intelligent Information Systems 16(3) (August 2001)
  25. Salamo, M., Golobardes, E.: Analysing rough sets weighting methods for case- based reasoning systems. Inteligencia Artificial. Revista Iberoamericana de IA (15) (2002) 10-18
  26. Salamó, M., Golobardes, E.: Rough sets reduction techniques for case-based rea- soning. In: Case-Based Reasoning Research and Development. Springer-Verlag (2001) 467-482
  27. Wierzbicki, J.: Rough set approach to CBR. In Ziarko, W., Yao, Y., eds.: Rough Sets and Current Trends in Computing. Volume 2005 of Lecture Notes in Artificial Intelligence., Berlin, Germany, Springer-Verlag (2001) 503-510
  28. Pawlak, Z.: Rough sets : Theoretical aspects of reasoning about data. Volume 9. Kluwer Academic Publishers, Dordrecht; Boston (1991) References
  29. Aha, David,. Klenk, Matthew., Laviers, Klenk ., Molineaux, David ,. Sukthankar, Gita. Opponent Modeling and Spatial Similarity to Retrieve and Reuse Superior Plays. In Proceedings of the ICCBR 2009 Workshops (2009) 97-106
  30. Mateas, Michael., Weber, Ben. Case-Based Reasoning for Build Order in Real- Time Strategy Games. In Proceedings of the Artificial Intelligence and Interactive Digital Entertainment (AIIDE 2009) [http://www.aaai.org/ocs/index.php/AIIDE/AIIDE09/paper/view/809/1084] (2009)
  31. Mateas, Michael., Weber, Ben. Conceptual Neighborhoods for Retrieval in Case- Based Reasoning. In Proceedings of the International Conference on Case-Based Reasoning (ICCBR 2009) (2009) 343-357
  32. Aamodt, Agnar., Szczepański, Tomasz. Case-Based Reasoning for Improved Mi- cromanagement in Real-Time Strategy Games. In Proceedings of the ICCBR 2009 Workshops (2009) 139-150
  33. Maslow, J., "Using Reinforcement Learning to Solve AI Control Problems."AI Game Programming Wisdom, Vol. 2 Charles River Media, 2003.
  34. Mateas, Michael., McCoy, Josh. An Integrated Agent for Playing Real-Time Strat- egy Games. In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008) (2008) 1313-1318
  35. Sharma, Manu., Holmes, Michael., Santamaria, Juan., Irani, Arya., Isbell, Charles L., Ram, Ashwin. Transfer Learning in Real-Time Strategy Games Using Hybrid CBR/RL. In Proceedings of the Twentieth International Joint Conference on Ar- tificial Intelligence (IJCAI) (2007) 1041-1046
  36. M. Buro. Real-time strategy games: A new AI research challenge. In IJCAI'2003, pages 1534-1535. Morgan Kaufmann, 2003.
  37. B. Díaz-Agudo, P. A. González-Calero, J. A. Recio-García, and A. A. Sánchez-Ruiz. Building CBR systems with jCOLIBRI. Special Issue on Experimental Software and Toolkits of the Journal Science of Computer Programming, 69(1-3):68-75, 2007.
  38. R. Fikes and N. J. Nilsson. Strips: A new approach to the application of theorem proving to problem solving. Artificial Intellicence, 2(3/4):189-208, 1971.
  39. M. Hall, E. Frank, G. Holmes, B. Pfahringer, P. Reutemann, and I. H. Witten. The WEKA data mining software: an update. ACM SIGKDD Explorations Newsletter, 11(1):10-18, June 2009.
  40. J. E. Laird and M. van Lent. Human-level AI's killer application: Interactive computer games. In AAAI 2000, pages 1171-1178, 2000.
  41. N. Llopis. Introduction to Game Development, chapter Game Architecture, pages 267-296. Charles River Media, 2005.
  42. M. Mateas and A. Stern. A behavior language for story-based believable agents. IEEE intel- ligent systems and their applications, 17(4):39-47, 2002.
  43. M. Molineaux and D. W. Aha. Tielt: A testbed for gaming environments. In AAAI, pages 1690-1691, 2005.
  44. S. Ontañón, K. Bonnette, P. Mahindrakar, M. Gómez-Martín, K. Long, J. Radhakrishnan, R. Shah, and A. Ram. Learning from human demonstrations for real-time case-based plan- ning. IJCAI-09 Workshop on Learning Structural Knowledge From Observations (STRUCK- 09), 2009.
  45. Ontañón, S., Mishra, K., Sugandh, N., Ram, A.: Case-based planning and execu- tion for real-time strategy games. In: 7th International Conference on Case-Based Reasoning. (2007) 164-178
  46. Mehta, M., Ontañón, S., Amundsen, T., Ram, A.: Authoring behaviors for games using learning from demonstration. In: Workshop on CBR for Computer Games at the 8th International Conference on Case-Based Reasoning. (2009)
  47. Thurau, C., Bauckhage, C.: Combining self organizing maps and multilayer per- ceptrons to learn bot-behavior for a commercial game. In: Proceedings of the GAME-ON Conference. (2003)
  48. Rubin, J., Watson, I.: SARTRE: System overview. A case-based agent for two- player texas hold'em. In: Workshop on CBR for Computer Games at the 8th International Conference on Case-Based Reasoning. (2009)
  49. Floyd, M.W., Esfandiari, B., Lam, K.: A case-based reasoning approach to imitat- ing RoboCup players. In: 21st International Florida Artificial Intelligence Research Society Conference. (2008) 251-256
  50. Floyd, M.W., Davoust, A., Esfandiari, B.: Considerations for real-time spatially- aware case-based reasoning: A case study in robotic soccer imitation. In: 9th European Conference on Case-Based Reasoning. (2008) 195-209
  51. Floyd, M.W., Esfandiari, B.: An active approach to automatic case generation. In: 8th International Conference on Case-Based Reasoning. (2009) 150-164
  52. Wooldridge, M.: An introduction to multiagent systems. John Wiley and Sons (2002)
  53. RoboCup: Robocup official site. http://www.robocup.org (2010)
  54. XPilot-AI: Xpilot-AI A library for writing XPilot bots. http://www.xpilot-ai.org (2010)
  55. Wendler, J., Lenz, M.: CBR for dynamic situation assessment in an agent-oriented setting. In: AAAI Workshop on Case-Based Reasoning Integrations. (1998)
  56. Steffens, T.: Adapting similarity measures to agent types in opponent modeling. In: AAMAS Workshop on Modelling Other Agents from Observations. (2004) 125-128
  57. Ros, R., Veloso, M., de Mántaras, R.L., Sierra, C., Arcos, J.L.: Retrieving and reusing game plays for robot soccer. In: 8th European Conference on Case-Based Reasoning. (2006) 47-61
  58. Aha, D.W., Molineaux, M., Sukthankar, G.: Case-based reasoning in transfer learning. In: 8th International Conference on Case-Based Reasoning. (2009) 29-44
  59. Molineaux, M., Aha, D.W., Moore, P.: Learning continuous action models in a real-time strategy environment. In: 21st International Florida Artificial Intelligence Research Society Conference. (2008) 257-262
  60. Auslander, B., Lee-Urban, S., Hogg, C., Muñoz-Avila, H.: Recognizing the en- emy: Combining reinforcement learning with strategy selection using case-based reasoning. In: 9th European Conference on Case-Based Reasoning. (2008) 59-73
  61. Romdhane, H., Lamontagne, L.: Reinforcement of local pattern cases for play- ing Tetris. In: 21st International Florida Artificial Intelligence Research Society Conference. (2008) 263-268
  62. Flinter, S., Keane, M.T.: On the automatic generation of cases libraries by chunking chess games. In: 1st International Conference on Case-Based Reasoning. (1995) 421-430
  63. Fagan, M., Cunningham, P.: Case-based plan recognition in computer games. In: 5th International Conference on Case-Based Reasoning. (2003) 161-170
  64. Atkeson, C.G., Schaal, S.: Robot learning from demonstration. In: Fourteenth International Conference on Machine Learning. (1997) 12-20
  65. Coates, A., Abbeel, P., Ng, A.Y.: Learning for control from multiple demonstra- tions. In: 25th International Conference on Machine Learning. (2008) 144-151
  66. Grollman, D.H., Jenkins, O.C.: Learning robot soccer skills from demonstration. In: IEEE International Conference on Development and Learning. (2007) 276-281
  67. Dinerstein, J., Egbert, P.K., Ventura, D., Goodrich, M.: Demonstration-based behavior programming for embodied virtual agents. Computational Intelligence 24(4) (2008) 235-256
  68. Craw, S. Agile case-based reasoning: A grand challenge towards opportunistic rea- soning from experiences. In Proceedings of the IJCAI-09 Workshop on Grand Chal- lenges in Reasoning from Experiences, pp. 33-39, Pasadena, CA, 2009.
  69. Leake, D.B and Kendall-Morwick, J. Four Heads Are Better than One: Combining Suggestions for Case Adaptation. In Proceedings of ICCBR-09, pp. 165-179.
  70. Zimmermann, A. Context-awareness in user modelling: Requirements analysis for a case-based reasoning application. In Case-Based Reasoning Research and Develop- ment, pp. 1064-1064, 2003.
  71. Chalmers, M. Abstract paths and contextually specific recommendations. In Pro- ceedings of DELOS/NSF Workshop on Personalisation and Recommender Systems in Digital Libraries, Dublin, June 2001.
  72. Ma, J. and Knight, B. A framework for historical case-based reasoning. In Case- Based Reasoning Research and Development, pp. 1067-1067, 2003.
  73. Champin, P.A., Briggs, P., Coyle, M. and Smyth, B. Coping with Noisy Search Experiences. In 29th SGAI International Conference on Artificial Intelligence (AI- 2009). Springer, pages 5-18, Cambridge, December 2009.
  74. Schwarz, S. and Roth-Berghofer, T. Towards goal elicitation by user observation. In Workshop on Knowledge and Experience Management at GI FGWM, LLWA 2003.
  75. Cordier, A., Mascret, B. and Mille, A. Extending Case-Based Reasoning with Traces. In Grand Challenges for reasoning from experiences, Workshop at IJCAI'09, Pasadena, CA. 2009.
  76. Georgeon, O., Henning, M. J., Bellet, T., and Mille, A. (2007). Creating Cognitive Models from Activity Analysis: A Knowledge Engineering Approach to Car Driver Modeling. International Conference on Cognitive Modeling, Ann Arbor, MI: Taylor & Francis, pp. 43-48.
  77. Leake, D.B.: Evaluating Explanations: A Content Theory. Lawrence Erlbaum Associates (1992)
  78. Roth-Berghofer, T.R., Cassens, J.: Mapping goals and kinds of explanations to the knowledge containers of case-based reasoning systems. In Muñoz-Avila, H., Ricci, F., eds.: Case Based Reasoning Research and Development -ICCBR 2005. Volume 3630 of LNAI., Chicago, Springer (2005) 451-464
  79. Leake, D., Whitehead, M.: Case provenance: The value of remembering case sources. Case-Based Reasoning Research and Development (2007) 194-208
  80. Aamodt, A., Plaza, E.: Case-based reasoning: Foundational issues, methodological variations, and system approaches. AI communications 7(1) (1994) 39-59
  81. Lockie, R.: Knowledge, Provenance and Psychological Explanation. Philosophy 79(03) (2004) 421-433
  82. Pearl, J.: Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann (1988)
  83. Jensen, F.V., Nielsen, T.D.: Bayesian Networks and Decision Graphs. 2nd edn. Springer (2007)
  84. Friedman, N., Goldszmidt, M.: Sequential update of bayesian network structure. In: Proceedings of the Thirteenth Conference on Uncertainty in Artificial Intelli- gence, Morgan Kaufmann (1997) 165-174
  85. Spiegelhalter, D.J., Lauritzen, S.L.: Sequential updating of conditional probabili- ties on directed graphical structures. Networks 20 (1990) 579-605
  86. Lacave, C., Díez, F.: A review of explanation methods for bayesian networks. The Knowledge Engineering Review 17(2) (2002) 107-127
  87. Aamodt, A., Langseth, H.: Integrating Bayesian networks into knowledge inten- sive CBR. In: American Association for Artificial Intelligence, Case-based reason- ing integrations; Papers from the AAAI workshop -Technical Report WS-98-15, Madison, WI., AAAI Press (1998) 1-6
  88. Sørmo, F., Cassens, J., Aamodt, A.: Explanation in case-based reasoning -per- spectives and goals. Artificial Intelligence Review 24(2) (October 2005) 109-143
  89. Richter, M.M.: The Knowledge Contained in Similarity Measures. Invited Talk at the First International Conference on Case-Based Reasoning, ICCBR'95, Sesimbra, Portugal (1995)
  90. Kofod-Petersen, A., Cassens, J.: Explanations and context in ambient intelligent systems. In kokinov, B., Richardson, D.C., Roth-Berhofer, T.R., Vieu, L., eds.: Modeling and Using Context, proceedings of the 6th International and Interdisci- plinary Conference (CONTEXT 2007). Volume 4635 of Lecture Notes in Artificial Intelligence., Roskilde, Denmark, Springer (August 2007) 303-316
  91. Simmhan, Y., Plale, B., Gannon, D.: A survey of data provenance in e-Science. SIGMOD Record 34(3) (2005) 31-36
  92. Leake, D., Whitehead, M.: Case provenance: The value of remembering case sources. In: Case-Based Reasoning Research and Development: Proceedings of the Seventh International Conference on Case-Based Reasoning, ICCBR-07, Berlin, Springer-Verlag (2007) 194-208
  93. Leake, D., Dial, S.: Using case provenance to propagate feedback to cases and adaptations. In: Proceedings of the Nineth European Conference on Case-Based Reasoning, Springer (2008) 255-268
  94. Briggs, P., Smyth, B.: Provenance, trust, and sharing in peer-to-peer case-based web search. In: Proceedings of the Nineth European Conference on Case-Based Reasoning, Springer (2008) 89-103
  95. Moreau, L.: The open provenance model. Technical report, University of Southampton (2008)
  96. Göker, M., Roth-Berghofer, T.: Development and utilization of a case-based help- desk support system in a corporate environment. In Althoff, K.D., Bergmann, R., Branting, L.K., eds.: Proceedings of the Third International Conference on Case-Based Reasoning, Springer Verlag (1999) 132-146
  97. Hammond, K.: Case-Based Planning: Viewing Planning as a Memory Task. Aca- demic Press, San Diego (1989)
  98. Veloso, M., Carbonell, J.: Derivational analogy in prodigy: Automating case ac- quisition, storage, and utilization. Machine Learning 10(3) (1993) 249-278
  99. Leake, D., Kinley, A., Wilson, D.: Learning to improve case adaptation by intro- spective reasoning and CBR. In: Proceedings of the First International Conference on Case-Based Reasoning, Berlin, Springer Verlag (1995) 229-240
  100. Leake, D., Powell, J.: Mining large-scale knowledge sources for case adaptation knowledge. In Weber, R., Richter, M., eds.: Proceedings of the Seventh Interna- tional Conference on Case-Based Reasoning, Berlin, Springer Verlag (2007) 209- 223
  101. Cox, M., Raja, A.: Metareasoning: A manifesto. Technical Memo 2028, BBN (2008)
  102. Arcos, J.L., Mulayim, O., Leake, D.: Using introspective reasoning to improve CBR system performance. In: Proceedings of the AAAI 2008 Workshop on Metar- easoning: Thinking About Thinking. (2008) 21-28
  103. Fox, S., Leake, D.: Modeling case-based planning for repairing reasoning failures. In: Proceedings of the 1995 AAAI Spring Symposium on Representing Mental States and Mechanisms, Menlo Park, CA, AAAI Press (March 1995) 31-38
  104. Simmhan, Y.L., Plale, B., Gannon, D.: Karma2: Provenance management for data driven workflows. International Journal of Web Services Research, Idea Group Publishing 5 (2008) 1-22
  105. Leake, D., Kendall-Morwick, J.: Towards case-based support for e-science work- flow generation by mining provenance information. In: Proceedings of the Nineth European Conference on Case-Based Reasoning, Springer (2008) 269-283
  106. Leake, D., Kendall-Morwick, J.: Four heads are better than one: Combining sugges- tions for case adaptation. In: Proceedings of the Eighth International Conference on Case-Based Reasoning, Springer (2009) 165-179
  107. Carbonell, J., Etzioni, O., Gil, Y., Joseph, R., Knoblock, C., Minton, S., Veloso, M.: Planning and learning in PRODIGY: Overview of an integrated architecture. In Ram, A., Leake, D., eds.: Goal-Driven Learning. MIT Press/Bradford Books, Cambridge, MA (1995) 297-305
  108. Moreau, L., Ludäscher, B., Altintas, I., Barga, R.S., Bowers, S., Callahan, S., Chin, Jr., G., Clifford, B., Cohen, S., Cohen-Boulakia, S., Davidson, S., Deelman, E., Digiampietri, L., Foster, I., Freire, J., Frew, J., Futrelle, J., Gibson, T., Gil, Y., Goble, C., Golbeck, J., Groth, P., Holland, D.A., Jiang, S., Kim, J., Koop, D., Krenek, A., McPhillips, T., Mehta, G., Miles, S., Metzger, D., Munroe, S., Myers, J., Plale, B., Podhorszki, N., Ratnakar, V., Santos, E., Scheidegger, C., Schuchardt, K., Seltzer, M., Simmhan, Y.L., Silva, C., Slaughter, P., Stephan, E., Stevens, R., Turi, D., Vo, H., Wilde, M., Zhao, J., Zhao, Y.: Special issue: The first provenance challenge. Concurr. Comput. : Pract. Exper. 20(5) (2008) 409-418
  109. Althoff, K.D., Aamodt, A.: Relating case-based problem solving and learning meth- ods to task and domain characteristics: Towards an analytic framework. Artificial Intelligence Communications 9 (1996) 11-6
  110. Aamodt, A.: Modeling the knowledge contents of CBR systems. In: Proceedings of the ICCBR 2001 Workshop on Case-Based Reasoning Authoring Support Tools. (2001) 32-37 Naval Research Lab Technical Note AIC-01-003.
  111. Birnbaum, L., Collins, G., Brand, M., Freed, M., Krulwich, B., Pryor, L.: A model- based approach to the construction of adaptive case-based planning systems. In Bareiss, R., ed.: Proceedings of the DARPA Case-Based Reasoning Workshop, San Mateo, Morgan Kaufmann (1991) 215-224
  112. Leake, D., Whitehead, M.: Case Provenance: The Value of Remembering Case Sources. In: Weber, R.O., Richter, M.M. (eds.) ICCBR 2007. LNCS (LNAI), vol. 4626, pp. 194- 208. Springer, Heidelberg (2007)
  113. Aha, D.W.: The Omnipresence of Case-Based Reasoning in Science and Application. Knowledge-Based Systems 11, 261-273 (1998)
  114. Leake, D., Dial, S.A.: Using Case Provenance to Propagate Feedback to Cases and Adap- tations. In: Althoff, K.-D. et al. (eds.) ECCBR 2008. LNCS (LNAI), vol. 5239, pp. 255- 268. Springer, Heidelberg (2008)
  115. Aamodt, A., Plaza, E.: Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches. Artificial Intelligence Communications 7, 39-59 (1994)
  116. López de Mántaras, R., McSherry, D., Bridge, D., Leake, D., Smyth, B., Craw, S., Falt- ings, B., Maher, M.L., Cox, M.T., Forbus, K., Keane, M., Aamodt, A., Watson, I.: Re- trieval, Reuse, Revision, and Retention in Case-Based Reasoning. Knowledge Engineer- ing Review 20, 215-240 (2005)
  117. McSherry, D.: Automating Case Selection in the Construction of a Case Library. Know- edge Based Systems 13, 133-140 (2000)
  118. Smyth, B., McKenna, E.: Competence Models and the Maintenance Problem. Computa- tional Intelligence 17, 235-249 (2001)
  119. Russell, S., Norvig, P.: Artificial Intelligence: A Modern Approach. Prentice Hall, Upper Saddle River (2010).
  120. McSherry, D: Demand-Driven Discovery of Adaptation Knowledge. In: 16th Interna- tional Joint Conference on Artificial Intelligence, pp. 222-227. Morgan Kaufmann, San Francisco (1999)
  121. Adrian, B., Hees, J., van Elst, L., Dengel, A.: iDocument: Using ontologies for extracting and annotating information from unstructured text. In: Mertsching, B., Hund, M., Aziz, Z. (eds.) KI 2009: Advances in Artificial Intelligence. Künstliche Intelligenz (KI-2009), Sept. 15-18, Paderborn, Germany. Lecture Notes in Artificial Intelligence, LNAI, vol. 5803, pp. 249-256. Springer, Heidelberg (2009)
  122. Bergmann, R., Schaaf, M.: Structural Case-Based Reasoning and Ontology-Based Knowledge Management: A Perfect Match? Journal of Universal Computer Science 9(7), 608-626 (2003), http://www.jucs.org/jucs_9_7/structural_case_based_ reasoning [Last access: 2010-02-26]
  123. Bizer, C., Heath, T., Berners-Lee, T.: Linked data -the story so far. Interna- tional Journal on Semantic Web and Information Systems 5(3), 1-22 (2009), http: //dblp.uni-trier.de/db/journals/ijswis/ijswis5.html#BizerHB09 [Last ac- cess: 2010-02-26]
  124. Buneman, P., Khanna, S., Tan, W.C.: Why and where: A characterization of data provenance. In: ICDT '01: Proceedings of the 8th International Conference on Database Theory. pp. 316-330. Springer, London, UK (2001)
  125. Cunningham, P., Doyle, D., Loughrey, J.: An evaluation of the usefulness of case- based explanation. Tech. Rep. TCD-CS-2003-04., Trinity College Dublin (2003)
  126. Green, T.J., Karvounarakis, G., Tannen, V.: Provenance semirings. In: PODS '07: Proceedings of the twenty-sixth ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems. pp. 31-40. ACM, New York, NY, USA (2007)
  127. Hartig, O.: Provenance Information in the Web of Data. In: Proceedings of the Linked Data on the Web LDOW Workshop at WWW (2009)
  128. Hellmann, S., Stadler, C., Lehmann, J., Auer, S.: Dbpedia live extraction. In: Debbabi, M., Caragea, D., Yang, X.C. (eds.) Proceedings of the 8th International Conference on Ontologies, DataBases, and Applications of Semantics (ODBASE). Lecture Notes in Computer Science, vol. 5871, pp. 1209-1223. Springer (2009)
  129. McGuinness, D.L., Ding, L., da Silva, P.P., Chang, C.: Pml 2: A modular expla- nation interlingua. In: Roth-Berghofer, T.R., Schulz, S., Leake, D.B. (eds.) ExaCt. pp. 49-55. AAAI Press (July 2007), technical Report WS-07-06
  130. McGuinness, D.L., Pinheiro da Silva, P.: Infrastructure for web explanations. In: Fensel, D., Sycara, K., Mylopoulos, J. (eds.) The Semantic Web -ISWC 2003. pp. 113-129 (2003)
  131. Newell, A.: The knowledge level. AI Magazine 2(2), 1-20 (1981)
  132. Richter, M.M.: The knowledge contained in similarity measures. Invited Talk at the First International Conference on Case-Based Reasoning, ICCBR'95, Sesimbra, Portugal (1995)
  133. Roth-Berghofer, T. (ed.): Künstliche Intelligenz: Anwendungen im Semantischen Web (Integriertes Seminar). Technische Universität Kaiserslautern (Juli 2004)
  134. Roth-Berghofer, T.: ExaCt manifesto: Explanation-aware computing (July 2009), http://on-explanation.net/content/ExaCt_Manifesto.html
  135. Roth-Berghofer, T., Adrian, B., Dengel, A.: Case acquisition from text: Ontology- based information extraction with SCOOBIE for myCBR. In: Montani, S., Bichin- daritz, I. (eds.) Advances in Case-Based Reasoning. Springer, Heidelberg (2010)
  136. Roth-Berghofer, T.R., Bahls, D.: Explanation capabilities of the open source case- based reasoning tool mycbr. In: Petridis, M., Wiratunga, N. (eds.) Proceedings of the thirteenth UK workshop on Case-Based Reasoning UKCBR 2008. pp. 23-34. University of Greenwich, London, UK (2008)
  137. Roth-Berghofer, T.R., Cassens, J.: Mapping goals and kinds of explanations to the knowledge containers of case-based reasoning systems. In: Muñoz-Avila, H., Ricci, F. (eds.) Case-Based Reasoning Research and Development, ICCBR 2005, Chicago, IL, USA, August 2005, Proceedings. pp. 451-464. No. 3620 in Lecture Notes in Artificial Intelligence LNAI, Springer, Heidelberg (2005)
  138. Stahl, A., Roth-Berghofer, T.R.: Rapid prototyping of CBR applications with the open source tool myCBR. In: Bergmann, R., Althoff, K.D. (eds.) Advances in Case-Based Reasoning. Springer (2008)
  139. Althoff, K.D., Bach, K., Deutsch, J.O., Hanft, A., Mänz, J., Müller, T., Newo, R., Reichle, M., Schaaf, M., Weis, K.H.: Collaborative multi-expert-systems -realizing knowlegde-product-lines with case factories and distributed learning systems. In: Baumeister, J., Seipel, D. (eds.) Workshop Proc. on the 3rd Workshop on Knowl- edge Engineering and Software Engineering (KESE 2007). Osnabrück (Sep 2007)
  140. Althoff, K.D., Hanft, A., Schaaf, M.: Case factory -maintaining experience to learn. In: Göker, M., Roth-Berghofer, T. (eds.) Proc. 8th European Conf. on Case- Based Reasoning (ECCBR'06), Ölüdeniz/Fethiye, Turkey. Springer, Berlin (2006)
  141. Bergmann, R.: Experience Management: Foundations, Development Methodology, and Internet-Based Applications, LNCS, vol. 2432. Springer (2002)
  142. Chakrabarti, S., Dom, B.E., Gibson, D., Kleinberg, J., Kumar, R., Raghavan, P., Rajagopalan, S., Tomkins, A.: Mining the link structure of the world wide web. IEEE Computer 32, 60-67 (1999)
  143. Church, K.W., Hanks, P.: Word association norms, mutual information, and lexi- cography. Computational Linguistics 16(1), 22-29 (1990)
  144. Cunningham, H., Maynard, D., Bontcheva, K., Tablan, V.: Gate: A framework and graphical development environment for robust nlp tools and applications. In: Proc. of the 40th Anniv.Meeting of the Assoc. for Comp. Linguistics (ACL'02) (2002)
  145. Fayyad, U., Piatetsky-Shapiro, G., Smyth, P.: The kdd process for extracting useful knowledge from volumes of data. Commun. ACM 39(11), 27-34 (1996)
  146. Feng, D., Shaw, E., Kim, J., Hovy, E.: An intelligent discussion-bot for answering student queries in threaded discussions. In: IUI '06: Proc. of the 11th Intl Conf. on Intelligent user interfaces. pp. 171-177. ACM Press, New York, NY, USA (2006)
  147. Ihle, N., Hanft, A., Althoff, K.D.: Extraction of adaptation knowledge from inter- net communities. In: Delany, S.J. (ed.) ICCBR 2009 Workshop Proc., Workshop Reasoning from Experiences on the Web. pp. 269-278 (July 2009)
  148. van der Linden, F., Schmid, K., Rommes, E.: Software Product Lines in Action - The Best Industrial Practice in Product Line Engineering. Springer, Berlin, Hei- delberg, Paris (2007)
  149. Liu, B.: Web Data Mining: Exploring Hyperlinks, Contents, and Usage Data. Springer-Verlag New York, Inc., Secaucus, NJ, USA (2008)
  150. Mika, P.: Flink: Semantic web technology for the extraction and analysis of social networks. Journal of Web Semantics 3, 211-223 (2005)
  151. Mika, P.: Social Networks and the Semantic Web, vol. 5. Springer Science+Business Media, LLC, Boston, MA (2007)
  152. Milne, P., Wiratunga, N., Lothian, R., Song, D.: Reuse of search experience for re- source transformation. In: Delany, S.J. (ed.) ICCBR 2009 Workshop Proc., Work- shop Reasoning from Experiences on the Web. pp. 45-54 (July 2009)
  153. Plaza, E., Baccigalupo, C.: Principle and praxis in the experience web: A case study in social music. In: Delany, S.J. (ed.) ICCBR 2009 Workshop Proc., Workshop Reasoning from Experiences on the Web. pp. 55-63 (July 2009)
  154. Reichle, M., Bach, K., Althoff, K.D.: The seasalt architecture and its realization within the docquery project. In: Mertsching, B. (ed.) Proc. of the 32nd Conference on Artificial Intelligence (KI-2009). pp. 556-563. LNCS (Sep 2009)
  155. Richter, M.M.: Introduction. In: Lenz, M., Bartsch-Spörl, B., Burkhard, H.D., Wess, S. (eds.) Case-Based Reasoning Technology -From Foundations to Appli- cations. LNAI 1400, Springer-Verlag, Berlin (1998)
  156. Sauer, C.S.: Analyse von Webcommunities und Extraktion von Wissen aus Com- munitydaten fr Case-Based Reasoning Systeme. Master's thesis, Institute of Com- puter Science, University of Hildesheim (2010)
  157. Smyth, B., Champin, P.A., Briggs, P., Coyle, M.: The case-based experience web. In: Delany, S.J. (ed.) ICCBR 2009 Workshop Proc., Workshop Reasoning from Experiences on the Web. pp. 74-82 (July 2009)
  158. Stahl, A., Roth-Berghofer, T.R.: Rapid prototyping of cbr applications with the open source tool mycbr. In: ECCBR '08: Proc. of the 9th European conference on Advances in Case-Based Reasoning. pp. 615-629. Springer, Heidelberg (2008) References
  159. Jill Freyne and Barry Smyth. Many Cases Make Light Work for Visualization in Many Eyes. In D. Bridge, E. Plaza, and N. Wiratunga, editors, Procs. of WebCBR: The Workshop on Reasoning from Experiences on the Web (at the 8th ICCBR), pages 25-44, 2009.
  160. Norman Ihle, Alexandre Hant, and Klaus-Dieter Althoff. Extraction of Adaptation Knowledge from Internet Communities. In D. Bridge, E. Plaza, and N. Wiratunga, editors, Procs. of WebCBR: The Workshop on Reasoning from Experiences on the Web (at the 8th ICCBR), pages 35-44, 2009.
  161. Richard Kittredge and John Lehrberger. Sublanguage: studies of language in re- stricted semantic domains. de Gruyter, 1982.
  162. Janet L. Kolodner. Case-Based Reasoning. Morgan Kaufmann, 1993.
  163. Enric Plaza. Semantics and Experience in the Future Web. In K.-D. Althoff, R. Bergmann, M. Minor, and A. Hanft, editors, Procs. of the 9th European Confer- ence on Case-Based Reasoning, LNCS 5239, pages 44-58. Springer Verlag, 2008.
  164. Enric Plaza and Claudio Baccigalupo. Principle and Praxis in the Experience Web: A Case Study in Social Music. In D. Bridge, E. Plaza, and N. Wiratunga, editors, Procs. of WebCBR: The Workshop on Reasoning from Experiences on the Web (at the 8th ICCBR), pages 55-63, 2009.
  165. Barry Smyth and Pierre-Antoine Champin. The Experience Web: A Case-Based Reasoning Perspective. In S. Craw, D. W. Aha, S. Singh, and B. Smyth, editors, Procs. of the Workshop on Grand Challenges for Reasoning From Experiences (at the 21st IJCAI), pages 53-61, 2009.
  166. Aidan Waugh and Derek Bridge. An Evaluation of the GhostWriter System for Case-Based Content Suggestions. In L. Coyle, J. Dunnion, and J. Freyne, editors, Procs. of the 20th Irish Conference on Artificial Intelligence and Cognitive Science, pages 264-273, 2009. References
  167. Kitano, H., Shimazu, H.: The Experience Sharing Architecture: A Case Study in Corporate-Wide Case-Based Software Quality Control. In, Case-Based Reasoning: Experiences, Lessons, & Future Directions. Leake, DB (Ed.) pp. 235-268 (1996)
  168. Hayes, C., Cunningham, P., Doyle, M.: Distributed CBR using XML. In: Proceedings of the KI-98 Workshop on Intelligent Systems and Electronic Commerce, Citeseer (1998)
  169. Shimazu, H.: A textual Case-Based Reasoning system using XML on the World- Wide Web. Advances in case-based reasoning (1998) 274-285
  170. Watson, I., Gardingen, D.: A distributed case-based reasoning application for engi- neering sales support. In: International Joint Conference on Artificial Intelligence. Volume 16., Citeseer (1999) 600-605
  171. Hayes, C., Cunningham, P.: Shaping a CBR view with XML. (Case-Based Reasoning Research and Development) 722-722
  172. Coyle, L., Doyle, D., Cunningham, P.: Representing Similarity for CBR in XML. (Advances in Case-Based Reasoning) 155-164
  173. Coyle, L., Hayes, C., Cunningham, P.: Representing cases for cbr in xml. Expert Update 6(2) (2003) 7-13
  174. Plaza, E., McGinty, L.: Distributed case-based reasoning. The Knowledge engineer- ing review 20(03) (2006) 261-265
  175. Berners-Lee, T., Hall, W., Hendler, J., O'Hara, K., Shadbolt, N., Weitzner, D.: A framework for Web Science. Foundations and Trends in Web Science 1(1) (2006) 76-77
  176. Bizer, C., Heath, T., Berners-Lee, T.: Linked Data-the story so far. Journal on Semantic Web and Information Systems (2009)
  177. Oren, E., Heitmann, B., Decker, S.: ActiveRDF: embedding Semantic Web data into object-oriented languages. Journal of Web Semantics (2008)
  178. Kolodner, J.: Retrieving events from a case memory: A parallel implementation. In: Proceedings of a Workshop on Case-Based Reasoning. Volume 233. (1988) 249
  179. Barletta, R., Mark, W.: Breaking cases into pieces. In: Proceedings of Case-Based Reasoning Workshop. (1988) 12-17
  180. Redmond, M.: Distributed cases for case-based reasoning: Facilitating use of multiple cases. In: Proceedings of AAAI. Volume 90. (1990) 304-309
  181. Chen, H., Wu, Z.: On Case-based Knowledge Sharing in Semantic Web. In: Inter- national Conference on Tools with Artificial Intelligence. (2003) 200-207
  182. Bichindaritz, I.: Mémoire: Case Based Reasoning Meets the Semantic Web in Biology and Medicine. In: European Case Based Reasoning Conference. (2004)
  183. d'Aquin, M., Lieber, J., Napoli, A.: Decentralized Case-Based reasoning for the Semantic Web. International Semantic Web Conference (2005) 142-155
  184. Decker, S., Melnik, S., Van Harmelen, F., Fensel, D., Klein, M., Broekstra, J., Erd- mann, M., Horrocks, I.: The semantic web: The roles of XML and RDF. IEEE Internet computing 4(5) (2000) 63-73
  185. Bojars, U., Passant, A., Cyganiak, R., Breslin, J.: Weaving SIOC into the Web of Linked Data. In: Linked Data on the Web Workshop. (2008)
  186. Richter, M.: The knowledge contained in similarity measures. Invited Talk at ICCBR 95 (1995)
  187. Hayes, C., Cunningham, P.: Shaping a CBR View with XML. In: International Conference on Case-Based Reasoning. (1999) 468-481
  188. Heitmann, B., Kinsella, S., Hayes, C., Decker, S.: Implementing semantic web ap- plications: reference architecture and challenges. In: Workshop on Semantic Web Enabled Software Engineering. (2009)
  189. Heitmann, B., Hayes, C.: Using linked data to build open, collaborative recommender systems. In: AAAI Spring Symposium: Linked Data Meets Artificial Intelligence'. (2010) References
  190. Balfe, E., Smyth, B.: A comparative analysis of query similarity metrics for community-based web search. In: ICCBR'05. pp. 63-77 (2005)
  191. Bridge, D., Plaza, E., Wiratunga, N. (eds.): Reasoning from Experiences on the Web. ICCBR, Seattle, WA (July 2009), ICCBR 2009 WORKSHOP
  192. Daniels, J.J., Rissland, E.L.: A case-based approach to intelligent information re- trieval. In: SIGIR'95. pp. 238-245 (1995)
  193. d'Aquin, M., Badra, F., Lafrogne, S., Lieber, J., Napoli, A., Szathmary, L.: Case base mining for adaptation knowledge acquisition. In: Proceedings of the Twentieth International Joint Conference on Artificial Intelligence (IJCAI-07). pp. 750-755. Morgan Kaufmann, San Mateo (2007)
  194. Freyne, J., Smyth, B.: Further experiments in case-based collaborative web search. In: ECCBR'06. pp. 256-270 (2006)
  195. Hendler, J.: Knowledge is power: A view from the semantic web. AI Magazine 26(4), 76-84 (2005)
  196. Ihle, N., Hanft, A., Althoff, K.D.: Extraction of adaptation knowledge from internet communities. In: WebCBR: Reasoning from Experiences on the Web (ICCBR 2009 Workshop) (2009)
  197. Leake, D., Powell, J.: Knowledge planning and learned personalization for Web- based case adaptation. In: Althoff, K., Bergmann, R., Minor, M., Hanft, A. (eds.) Advances in Case Based Reasoning: 9th European Conference. pp. 284-298. Springer Verlag, Berlin (2008)
  198. Leake, D., Powell, J.: A general introspective reasoning approach to web search for case adaptation. In: Bichindaritz, I., Montani, S. (eds.) Proceedings of the Ninth International Conference on Case-Based Reasoning. Springer Verlag, Berlin (2010), in press
  199. Mantaras, R., McSherry, D., Bridge, D., Leake, D., Smyth, B., Craw, S., Faltings, B., Maher, M., Cox, M., Forbus, K., Keane, M., Aamodt, A., Watson, I.: Retrieval, reuse, revision, and retention in CBR. Knowledge Engineering Review 20(3) (2005)
  200. Plaza, E., Baccigalupo, C.: Principle and praxis in the experience web: A case study in social music. In: WebCBR: Reasoning from Experiences on the Web (ICCBR 2009 Workshop) (2009)
  201. Smyth, B., McClave, P.: Similarity vs. diversity. In: Case-Based Reasoning Re- search and Development: Proceedings of the Fourth International Conference on Case-Based Reasoning, ICCBR-01. pp. 347-361. Springer-Verlag, Berlin (2001) References
  202. C. Buckley, G. Salton, J. Allan, and A. Singhal. Automatic query expansion using smart: Trec 3. In TREC, pages 0-, 1994.
  203. O. G. F. Diaz, R. S. Salgado, I. S. Moreno, and G. Rodrguez-Ortiz. Searching and selecting web services using case based reasoning. In ICCSA (4), Lecture Notes in Computer Science, pages 50-57. Springer, 2006.
  204. J. J. Gibson. The theory of affordances. In R. Shaw and J. Bransford, editors, In Perceiving, Acting, and Knowing, 1977.
  205. S. Ha. A personalized counseling system using case-based reasoning with neural symbolic feature weighting (cansy). Appl. Intell., 29(3):279-288, 2008.
  206. E. Hatcher and O. Gospodnetic. Lucene in Action (In Action series). Manning Publications Co., Greenwich, CT, USA, 2004.
  207. B. Limthanmaphon and Y. Zhang. Web service composition with case-based reasoning. In ADC '03: Proceedings of the 14th Australasian database conference, pages 201-208, Darlinghurst, Australia, Australia, 2003. Australian Computer Society, Inc.
  208. P. Majumder, M. Mitra, S. Parui, and P. Bhattacharyya. Initiative for indian language ir evaluation. In The First International Workshop on Evaluating Information Access, pages 14-16, Tokyo, Japan, 2007. EVIA.
  209. A. Singhal, G. Salton, M. Mitra, and C. Buckley. Document length normalization. Inf. Process. Manage., 32(5):619-633, 1996.
  210. R. Cross, R. E. Rice, and A. Parker. Information seeking in social context: structural influ- ences and receipt of information benefits. IEEE Transactions on Systems, Man, and Cyber- netics, Part C, 31(4):438-448, 2001.
  211. B. M. Evans and E. H. Chi. Towards a model of understanding social search. In CSCW '08: Proceedings of the ACM 2008 conference on Computer supported cooperative work, pages 485-494, New York, NY, USA, 2008. ACM.
  212. E. A. Fox, D. Hix, L. T. Nowell, D. J. Brueni, D. Rao, W. C. Wake, and L. S. Heath. Users, user interfaces, and objects: Envision, a digital library. J. Am. Soc. Inf. Sci., 44(8):480-491, 1993.
  213. D. Horowitz and S. D. Kamvar. The anatomy of a large-scale social search engine. In WWW, 2010.
  214. I. Karasavvidis. Distributed Cognition and Educational Practice. Journal of Interactive Learning Research, pages 11-29, 2002.
  215. J. D. Lafferty, A. McCallum, and F. C. N. Pereira. Conditional random fields: Probabilistic models for segmenting and labeling sequence data. In Proceedings of the Eighteenth Inter- national Conference on Machine Learning, ICML '01, pages 282-289, San Francisco, CA, USA, 2001. Morgan Kaufmann Publishers Inc.
  216. K. McCarthy, L. McGinty, B. Smyth, and M. Salamó. The needs of the many: A case-based group recommender system. Advances in Case-Based Reasoning, 4106:196-210, 2006.
  217. M. J. Pazzani and D. Billsus. Content-based recommendation systems. pages 325-341, 2007.
  218. S. E. Robertson and S. Walker. Some simple effective approximations to the 2-poisson model for probabilistic weighted retrieval. In Proceedings of the 17th annual international ACM SIGIR conference on Research and development in information retrieval, SIGIR '94, pages 232-241, New York, NY, USA, 1994. Springer-Verlag New York, Inc.
  219. B. Rogoff. Apprenticeship in thinking: Cognitive development in social context. Oxford University Press New York, 1990.
  220. S. Sahay, S. Ahn, S.-C. Lu, B. Sherwell, A. Venkatesh, and A. Ram. Healthbuzz: Contextual social search and conversations. In The Third Annual Workshop on Search in Social Media (SSM 2010), February 2010.
  221. S. Sahay, S. Mukherjea, E. Agichtein, E. V. Garcia, S. B. Navathe, and A. Ram. Discovering semantic biomedical relations utilizing the web. ACM Trans. Knowl. Discov. Data, 2(1):1- 15, 2008.
  222. S. Sahay, A. Venkatesh, and A. Ram. Cobot: Real time multi user conversational search and recommendations. In Recommender Systems and the Social Web, volume 532. CEUR Workshop Proceedings, 2009.
  223. B. Smyth, P. Briggs, M. Coyle, and M. P. O'Mahony. A case-based perspective on social web search. In Proceedings of the 8th International Conference on Case-Based Reason- ing: Case-Based Reasoning Research and Development, pages 494-508, Berlin, Heidelberg, 2009. Springer-Verlag.
  224. B. Wilson and H. Meij. Constructivist learning environments: Case studies in instructional design. IEEE Transactions on Professional Communication, pages 0361-1434, 1997.
  225. O. Ybarra, E. Burnstein, P. Winkielman, M. C. Keller, M. Manis, E. Chan, and J. Rodriguez. Mental Exercising Through Simple Socializing: Social Interaction Promotes General Cogni- tive Functioning. Pers Soc Psychol Bull, 34(2):248-259, 2008.
  226. Herrera, P.J., Iglesias, P., Romero, D., Rubio, I., Díaz-Agudo, B.: Jadacook: Java application developed and cooked over ontological knowledge. In Schaaf, M., ed.: ECCBR Workshops. (2008) 209-218
  227. Herrera, P.J., Iglesias, P., Sánchez, A.M.G., Díaz-Agudo, B.: Jadacook 2: Cooking over ontological knowledge. In: ICCBR Workshops. Computer Cooking Contest. (2009)
  228. Recio-García, J.A., D'iaz-Agudo, B., González-Calero, P.A.: Boosting the perfor- mance of cbr applications with jcolibri. In: ICTAI, IEEE Computer Society (2009) 276-283
  229. Fields, D.K., Kolb, M.A., Bayern, S.: Web Development with Java Server Pages. Manning Publications Co., Greenwich, CT, USA (2001)
  230. Buchholz, S., Marsi, E.: Conll-x shared task on multilingual dependency pars- ing. In: Proceedings of the 10th Conference on Computational Natural Language Learning (CoNLL-X). (2006) 149-164
  231. Fellbaum, C., ed.: WordNet: an electronic lexical database. MIT Press (1998)
  232. Hanft, A., Ihle, N., Bach, K., Newo, R.: Cookiis -competing in the first computer cooking contest. KI 23 (2009) 30-33
  233. Lin, D.: Dependency-based evaluation of minipar. In: Proc. Workshop on the Evaluation of Parsing Systems. Granada (1998)
  234. Nivre, J., Hall, J., Nilsson, J.: Maltparser: A data-driven parser-generator for dependency parsing. In: In Proc. of LREC-2006. (2006) 2216-2219
  235. Sampson, G.: (Briefly noted english for the computer: The susanne corpus and analytic scheme) References
  236. Badra, F., Bendaoud, R., Bentebitel, R., Champin, P.A., Cojan, J., Cordier, A., Després, S., Jean-Daubias, S., Lieber, J., Meilender, T., Mille, A., Nauer, E., Napoli, A., Toussaint, Y.: Taaable: Text Mining, Ontology Engineering, and Hierarchical Classification for Tex- tual Case-Based Cooking. In: ECCBR Workshops, Workshop of the First Computer Cooking Contest. (2008) 219-228
  237. Badra, F., Cojan, J., Cordier, A., Lieber, J., Meilender, T., Mille, A., Molli, P., Nauer, E., Napoli, A., Skaf-Molli, H., Toussaint, Y.: Knowledge Acquisition and Discovery for the Textual Case-Based Cooking system WikiTaaable. In: Workshops of the 8th International Conference on Case-Based Reasoning (ICCBR-09). (2009)
  238. Cojan, J., Lieber, J.: Belief Merging-based Case Combination. In David C. Wilson, Lorraine McGinty, eds.: 8th International Conference on Case-Based Reasoning -ICCBR 2009 Case- Based Reasoning Research and Development. Volume 5650 of Lecture Notes in Computer Science., Seattle United States, Springer Berlin (07 2009) 105-119 The original publication is available at www.springerlink.com.
  239. Dufour-Lussier, V., Lieber, J., Nauer, E., Toussaint, Y.: Text adaptation using formal concept analysis. In Bichindaritz, I., Montani, S., eds.: Case-Based Reasoning Research and Devel- opment. Volume 6176 of Lecture Notes in Artificial Intelligence., Alessandria, Italy, Springer (2010) 96-110
  240. Alchourrón, C.E., Gärdenfors, P., Makinson, D.: On the logic of theory change: partial meet functions for contraction and revision. Journal of Symbolic Logic 50 (1985) 510-530
  241. Asiimwe, S., Craw, S., Wiratunga, N., Taylor, B.: Automatically acquiring structured case representations: The SMART way. In: Applications and Innovations in Intelligent Systems XV, Springer (2007)
  242. Carpineto, C., Romamo, G.: Order-Theoretical Ranking. Journal of the American Society for Information Science 51(7) (2000)
  243. Bendaoud, R., Napoli, A., Toussaint, Y.: Formal Concept Analysis: A unified framework for building and refining ontologies. In Aldo Gangemi, Jérôme Euzenat, eds.: 16th International Conference on Knowledge Engineering and Knowledge Management -EKAW 2008. Volume 5268., Acitrezza, Catania Italie, Springer Berlin / Heidelberg (2008) 156-171
  244. Ganter, B., Wille, R.: Formal Concept Analysis. Springer, Heidelberg (1999) References
  245. Ihle, N., Newo, R., Hanft, A., Bach, K., Reichle, M.: Cookiis -A Case-Based Recipe Advisor. In Delany, S.J., ed.: Workshop Proceedings of the 8th International Conference on Case-Based Reasoning, Seattle, WA, USA (July 2009) 269-278
  246. Hanft, A., Newo, R., Bach, K., Ihle, N., Althoff, K.D.: Cookiis -a successful recipe advisor and menu advisor. In Montani, S., Jain, L., eds.: Successful Case-based Reasoning applications. Springer (2010)
  247. Bergmann, R.: Experience Management: Foundations, Development Methodology, and Internet-Based Applications. Volume 2432 of LNAI. Springer-Verlag (2002)
  248. Badra, F., Cojan, J., Cordier, A., Lieber, J., Meilender, T., Mille, A., Molli, P., Nauer, E., Napoli, A., Skaf-Molli, H., Toussaint, Y.: Knowledge acquisition and discovery for the textual case-based cooking system WikiTaaable. In Delany, S.J., ed.: ICCBR 2009, Workshop Proc. (2009)
  249. Herrera, P.J., Iglesias, P., Sanchez, A.M.G., Diaz-Agudo, B.: Jadacook 2: Cooking over ontological knowledge. In Delany, S.J., ed.: ICCBR 2009, Workshop Proc. (2009)
  250. Ihle, N., Hanft, A., Althoff, K.D.: Extraction of adaptation knowledge from internet communities. In Delany, S.J., ed.: ICCBR 2009, Workshop Proc. (2009) 35-44 References
  251. F. Badra, J. Cojan, A. Cordier, J. Lieber, T. Meilender, A. Millle, P. Molli, E. Nauer, A. Napoli, H. Skaf-Molli, and Y. Toussaint. "Knowledge Acquisition and Discovery for the Textual Case-Based Cooking System WIKIAAABLE". 8 th International Conference on Case-Based Reasoning -ICCBR 2009, Workshop Proceedings 2009, Page 249 -258.
  252. K. J. Hammond. "CHEF: A model of case-based planning." AAAI Proceedings of the 5 th National Conference on Artificial Intelligence, page 267-271. Morgan Kaufmann, 1986.
  253. S. T. Herbst, "Food Lover's Companion, The (Barron's Cooking Guide) 2 nd Edition", available at: http://www.epicurious.com/tools/fooddictionary.
  254. N. Ihle, R. Newo, A. Hanft, K. Bach, M. Reichle. "CookIIS -A Case-Based Recipe Advisor". 8 th International Conference on Case-Based Reasoning - ICCBR 2009, Workshop Proceedings 2009.
  255. D. Leake. "Case-Based Reasoning: Experience, Lessons, and Future Directions". Menlo Park: AAAI Press/MIT Press, 1996.
  256. K. H. Xu, J. Zhang, S. Gao. "High-Order Functions and their Ordering Relations". The Fifth International Conference on Digital Information Management, 2010.
  257. K. H. Xu, J. Zhang, S. Gao. "Froglingo, A Monolithic Alternative to DBMS, Programming Language, Web Server, and File System". The Fifth International Conference on Evaluation of Novel Approaches to Software Engineering, 2010.
  258. K. H. Xu, J. Zhang, S. Gao. "An Assessment on the Easiness of Computer Languages". To appear in the Journal of Information Technology Review, 2010.
  259. K. H. Xu, J. Zhang, S. Gao, R. R. McKeown. "Let a Data Model be a Class of Total Recursive Functions". The 2010 International Conference on Theoretical and Mathematical Foundations of Computer Science (TMFCS-10), 2010.
  260. K. H. Xu, J. Zhang. "A User's Guide to Froglingo, An Alternative to DBMS, Programming Language, Web Server, and File System, Release 1.0, January 2010". Available at the website: http://www.froglingo.com/ FrogUserGuide10.doc.
  261. K. H. Xu, J. Zhang, S. Gao. "Assessing Easiness with Froglingo". The Second International Conference on the Application of Digital Information and Web Technologies, 2009, page 847 -849.

Related papers

Cobot: Real Time Multi User Conversational Search and Recommendations
Saurav Sahay

2009

ABSTRACT Cobot is a new intelligent agent platform that connects users through real-time and off-line conversations about their health and medical issues. Intelligent web based information agents (conversational/community bots) participate in each conversation providing highly-relevant real-time informational recommendations and connecting people with relevant conversations and other community members.

View PDFchevron_right
Conversation Strategy of a Chatbot for Interactive Recommendations
Hung-Hsuan Huang

Intelligent Information and Database Systems, 2018

This paper presents a conversation strategy for interactive recommendations using a chatbot. Chatbots are attracting attention to provide a flexible user interface using natural language for various domains. For a given task, what kind of questions to ask and/or what information should be provided and how to process user responses play a crucial role in developing an effective chatbot. In this paper, we focus on a task of recommending an item that suits a user's preference and propose a conversation strategy where a chatbot combines questions about user's preference and recommendations soliciting user's feedback to them. The balance between questions and recommendations is controlled by changing the parameter values. We target a chatbot that uses a graphical user interface (GUI) and apply approaches proposed in the field of recommendation systems. Preliminary experiment results with a prototype indicate the potential of our proposed approach.

View PDFchevron_right
A Framework for Building Chat-based Recommender Systems
Giovanni Semeraro

2018

Chat-based recommender systems are getting more and more attention in recent time given their natural interaction with the user. Indeed, chat-based recommender systems implement a paradigm where users define their preferences and discover items that best fit their needs through a dialog. A chat-based recommender system can be easily integrated in platforms such as social networks, e-commerce websites, bank websites. Therefore, the preferences can be directly provided by the users during the dialog or can be automatically extracted from their activities on the same platform that hosts the chatbot [3]. In this demo, we present a framework for building chatbased recommender systems. The framework, based on a content-based recommendation algorithm, is independent from the domain.

View PDFchevron_right
IrisBot : An Open-Domain Conversational Bot for Personalized Information Access
Harshita Sahijwani

2018

We describe IrisBot, a conversational agent that aims to help a customer be informed about the world around them, while being entertained and engaged. Our bot attempts to incorporate real-time search, informed advice, and latest news recommendation into a coherent conversation. IrisBot can already track information on the latest topics and opinions from News, Sports, and Entertainment and some specialized domains. The key technical innovations of IrisBot are novel algorithms for contextualized classification of the topic and intent of the user’s utterances, modular ranking of potential responses, and personalized topic suggestions. Our preliminary experimental results based on overall customer experience ratings and A/B testing analysis, focus on understanding the contribution of both algorithmic and surface presentation features. We also suggest promising directions for continued research, primarily focusing on increasing the coverage of topics for in-depth domain understanding, fu...

View PDFchevron_right
HALPIN: a multimodal and conversational system for information seeking on the World Wide Web
Jean Caelen

… on Accessing Information in Spoken Audio, 1999

Giving to computers the ability to talk and understand a natural language conversation is a major field of research. We have developed the HALPIN (Hyperdialogue avec un Agent en Langage Proche de l&#x27;Interaction Naturelle) system to implement our multimodal conversational model for information retrieval. This dialogue-oriented interface allows the access to the INRIA&#x27;s database (Institut National de Recherche en Informatique et Automatique, 83297 documents available), on the internet, in a natural language (NL) way ...

View PDFchevron_right
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved