Explanation aware computing

Web-based science analysis and processing tools allow users to access, analyze, and generate visualizations of data without requiring the user be an expert in data processing. These tools simplify science analysis for all science users by reducing the data processing overhead for the user. The benefits of these tools come with a cost, the increased need for transparency in data processing. By providing a clear explanation of the science concepts and processing performed by the science analysis tool we can increase user trust, understanding, and accountability and reduce misinterpretation or generation of inconsistent results. We will demonstrate knowledge provenance (processing lineage and related domain information) presentation capabilities applied to an existing web-based Earth science data analysis tool (e.g. Giovanni from NASA/GSFC). Our conclusion is that user accessible visual presentations of knowledge provenance are key to building meaningful user understanding of analysis and processing decisions and should be a key component of data analysis tools. 1 Data access, format translations, data calibrations and screenings, etc.

Scientific data services are increasing in usage and scope, and with these increases comes growing need for access to provenance information. Our goal is to design and implement an extensible provenance solution that is deployed at the science data ingest time. In this paper, we describe our work in the setting of a particular set of data services in the area of solar coronal physics. The paper focuses on one existing federated data service and one proposed observatory. Our claim is both that the design and ...

As personal assistant software matures and assumes more autonomous control of its users' activities, it becomes more critical that this software can explain its task processing. It must be able to tell the user why it is doing what it is doing, and instill trust in the user that its task knowledge reflects standard practice and is being appropriately applied. We will describe the ICEE (Integrated Cognitive Explanation Environment) explanation system and its approach to explaining task reasoning. Key features include (1) an architecture ...

Information retrieval and integration systems typically must handle incomplete and inconsistent data. Current approaches attempt to reconcile discrepant information by leveraging data quality, user preferences, or source provenance information. Such approaches may overlook the fact that information is interpreted relative to its context. Therefore, discrepancies may be explained and thereby resolved if contexts are taking into account. In this paper, we describe an information integrator that is capable of explaining ...

In order for agents and humans to leverage the growing wealth of heterogeneous information and services on the web, increasingly, they need to understand the information that is delivered to them. In the simplest case, an agent or human is retrieving ���look-up��� information and would benefit from having access to provenance information concerning recency, source authoritativeness, etc. In more complicated situations where information is manipulated before it is returned as an answer, agents and humans would benefit from ...

The main advantage of machine learning algorithms that learn simple symbolic models is in their capability to trivially provide justifications for their decisions. However, there is no guarantee that these justifications will be understood by experts and other users. Induced models are often strange to the domain experts as they understand the problem in a different way. We suggest the use of argument based machine learning (ABML) to deal with this problem. This approach combines machine learning with explanations provided by domain experts. An ABML method is required to learn a model that correctly predicts learning examples and is consistent with the provided explanations. The present paper describes an application of ABML to learning a complex chess concept of an attack on the castled king. The explanation power of the learned model for this concept is especially important, as it will be used in a chess tutoring application.

Abstract. Across many fields involving complex computing, software systems are being augmented with workflow logging functionality. The log data can be effectively organized using declarative structured languages such as OWL; however, such declarative encodings alone are not enough to facilitate understandable workflow systems with high quality explanation. In this paper, we present our approach for visually explaining OWL-encoded workflow logs for complex systems, which includes the following steps:(i) identifying and ...

The main advantage of machine learning algorithms that learn simple symbolic models is in their capability to trivially provide justifications for their decisions. However, there is no guarantee that these justifications will be understood by experts and other users. Induced models are often strange to the domain experts as they understand the problem in a different way. We suggest the use of argument based machine learning (ABML) to deal with this problem. This approach combines machine learning with explanations provided by domain experts. An ABML method is required to learn a model that correctly predicts learning examples and is consistent with the provided explanations. The present paper describes an application of ABML to learning a complex chess concept of an attack on the castled king. The explanation power of the learned model for this concept is especially important, as it will be used in a chess tutoring application.

The number of news reports published online is now so great that it is impossible for any person to read all of them. Not all of these reports are equally interesting. Automating the identification and evaluation of interest in news is therefore a potentially valuable goal. Traditional methods of information management such as information filtering, information retrieval and collaborative filtering are unsuited to the task of filtering news. This abstract presents a framework that permits the identification of interesting news by means of the identification of violated expectations. Facts derived from news reports, expectations and related background knowledge can be used to (i) justify the decision to rate news as interesting, (ii) explain why the information in the report is unexpected and, (iii) explain the context in which the report appears.

Large volumes of news are available around the clock and around the world. It is increasingly difficult to sort interesting from uninteresting news. Information filtering and retrieval have addressed this problem by comparing lexical information in documents with a query or profile of user's interests. However, this approach does little to address the fact that the most interesting news is usually that which is unexpected. This thesis presents a framework that addresses this problem by reasoning about the information in reports, comparing it with the general knowledge of background information and expectations. This framework makes it possible to identify and rank unexpected, and therefore interesting, news. The Expectation Violation Analysis (EVA) framework presented here draws reasoned conclusions as to why the information in a report is or is not interesting. The EVA framework contains a set of background knowledge and a set of expectations which represent the usual state of ...

Introduction. Social Semantic software, for example Semantik Wikis, combines collaborative work and social interaction between users with Semantic Web technologies. Reasoning, seen as the generation of new data from declarative specifications, and querying, seen as high-level and declarative access and data construction, are distinguishing features of Semantic software. In order to benefit non-expert users of social software, both technologies should be easy to use while at the same time providing expressive power to advanced users. ...

Case-based reasoning (CBR) is widely applicable to the diagnosis of problems and the identification of solutions to them. This review of the literature identifies key papers relating to this use of CBR. The stages in diagnosing and troubleshooting a problem are considered and the elements of a CBR system for achieving this are described. Five systems are discussed that have frequently been cited in the literature and illustrate the use of CBR in diagnosis and troubleshooting applications. 2 Elements of a case-based diagnosis application Diagnosing and troubleshooting a problem typically involves three stages.

In order to assist a power plant operator to face un- usual situations, we have developed an intelligent as- sistant that explains the suggested commands generated by an MDP-based planning system. This assistant pro- vides the trainee a better understanding of the recom- mended actions to later generalize them to similar situ- ations. In a first stage, built-in explanations are

In this paper we describe the application of a novel learning and problem solving architecture to the domain of airspace management, where multiple requests for the use of airspace need to be reconciled and managed automatically. The key feature of our “Generalized Integrated Learning Architecture”(GILA) is a set of integrated learning and reasoning (ILR) systems coordinated by a central meta-reasoning executive (MRE). Each ILR learns independently from the same training example and contributes to problem- ...

Abstract. One current challenge in linked science is to adequately describe where a piece of information in the linked science cloud came from. Provenance models, such as Proof Markup Language (PML), have developed methods for expressing simple relationships between information and the sources of information. We argue that the representation of where information comes from is central to trusting linked data in scientific applications. We introduce the notion of a model of information source and the usage of the source to ...

Many abductive understanding systems explain novel situations by a c haining process that is neutral to explainer needs beyond generating some plausible explanation for the event being explained. This paper examines the relationship of standard models of abductive understanding to the case-based explanation model. In case-based explanation, construction and selection of abductive h ypotheses are focused by speci c explanations of prior episodes and by goal-based criteria re ecting current information needs. The case-based method is inspired by observations of human explanation of anomalous events during everyday understanding, and this paper focuses on the method's contributions to the problems of building good explanations in everyday d o m a i n s. W e i d e n tify ve central issues, compare how those issues are addressed in traditional and case-based explanation models, and discuss motivations for using the case-based approach to facilitate generation of plausible and useful explanations in domains that are complex and imperfectly understood.

Explanation and argumentation are fundamental to reasoning. They are therefore of some importance to artificial intelligence. Discourse-based reasoning (DBR) is a knowledge representation technology that uses natural patterns of discourse as a basis for a structural ontology of explanatory and argumentative reasoning. By this means, we may ontologize the reasoning process itself, thus rendering it accessible as an explanatory mechanism. Towards this objective, this paper introduces three general categories of rhetorical relations, including inferential, synthetic, and multinuclear. Inferential relations are argumentative, causal, or conditional; synthetic relations are purely explanatory; and multinuclear relations are used to express rhetorically bound pluralities of concept instantiations. These categories are used to explore design-time and runtime dimensions for representing linked and convergent structures. Using discourse-based reasoning, human and artificial agents will engage in collaborative reasoning, discover knowledge, resolve conflict, and render explanations using rhetorically explicit representations.

Numerous studies have affirmed the value of asynchronous online communication as a learning resource. Several investigations, however, have indicated that discussions in asynchronous environments are often neither interactive nor coherent. This research sought to develop an enhanced understanding of interactional coherence in asynchronous learning environments. The study used Rhetorical Structure Theory (RST) to analyze and assess the coherence of a several asynchronous discussions. The analysis revealed that the discussions were structurally dynamic. While RST structures resulting from static documents are acyclic tree-shaped structures, the rhetorical networks representing asynchronous threads are frequently cyclic. Thus, the analysis required a modified form of RST based on reduced constraints and restricted schemas. By this means, it was possible to create structural models of the discussions. These models were used to investigate asynchronous argumentation and topic drift and to perform a comparative analysis of multiple discussions. I wish to express my deepest appreciation to the many people to have made this dissertation possible. I am especially grateful to the members of my dissertation committee. Dr. Laurie Dringus recognized before I did that the vague ideas I was muddling over were potential dissertation material. It was only with her continuous support and encouragement that I was able, slowly and sometimes not so surely, to transform those ideas into this report. Dr. Steven Zink has shown me a new standard of excellence in research and expository prose; and I have sought, as best I can, to emulate his approach. Dr. Marlyn Littman's careful eye has helped me avoid egregious errors more than once during my research and writing. Together these committee members have worked as a team, shepherding me through the dissertation process far more smoothly than I could have imagined.

Abstract* This paper describes the inference explanation capabilities of Cyc, a logical reasoning system that includes a huge “com-monsense” knowledge base and an inference engine that sup-ports both question answering and hypothesis generation. Cyc allows the user to ...

We present an overview of different theories of explanation from the philosophy and cognitive science communities. Based on these theories, as well as models of explanation from the knowledge-based systems area, we present a framework for explanation in case-based reasoning (CBR) based on explanation goals. We propose ways that the goals of the user and system designer should be taken into account when deciding what is a good explanation for a given CBR system. Some general types of goals relevant to many CBR systems are identified, and used to survey existing methods of explanation in CBR. Finally, we identify some future challenges.

Artificial Intelligence techniques are increasingly being used to develop smart training applications for professionals in various domains. This paper presents an intelligent training system that enables professionals in the public domain to practice their aggression de-escalation skills. The system is one of the main products of the STRESS project, an interdisciplinary research project involving partners from academia, industry and society. The system makes use of a variety of AI-related techniques, including simulation, virtual agents, sensor fusion, model-based analysis and adaptive support. A preliminary evaluation of the system has been conducted with two groups of potential end users, namely tram conductors and police academy students.

In this paper we describe the application of a novel learning and problem solving architecture to the domain of airspace management, where multiple requests for the use of airspace need to be reconciled and managed automatically. The key feature of our “Generalized Integrated Learning Architecture”(GILA) is a set of integrated learning and reasoning (ILR) systems coordinated by a central meta-reasoning executive (MRE). Each ILR learns independently from the same training example and contributes to problem- ...

UNDERSTANDING WHAT MAY HAVE HAPPENED IN DYNAMIC, PARTIALLY OBSERVABLE ENVIRONMENTS Matthew Molineaux, PhD George Mason University, 2017 Dissertation Director: Dr. Gheorghe Tecuci In this work, we address the problem of understanding what may have happened in a goal-based deliberative agent’s environment after the occurrence of exogenous actions and events. Such an agent observes, periodically, information about the state of the world, but this information is incomplete, and reasons for state changes are not observed. We propose methods a goal-based agent can use to construct internal, causal explanations of its observations based on a model of its environment. These explanations comprise a series of inferred actions and events that have occurred and continue to occur in its world, as well as assumptions about the initial state of the world. We show that an agent can more accurately predict future events and states by reference to these explanations, and thereby more reliably achieve...

In this paper we describe the application of a novel learning and problem solving architecture to the domain of airspace management, where multiple requests for the use of airspace need to be reconciled and managed automatically. The key feature of our “Generalized Integrated Learning Architecture”(GILA) is a set of integrated learning and reasoning (ILR) systems coordinated by a central meta-reasoning executive (MRE). Each ILR learns independently from the same training example and contributes to problem- ...

In the past five years, we have designed and evolved an interlingua for sharing explanations generated by various automated systems such as hybrid web-based question answering systems, text analytics, theorem proving, task processing, web services execution, rule engines, and machine learning components. In this paper, we present our recent major updates including:(i) splitting the interlingua into three modules (ie provenance, information manipulation or justifications, and trust) to reduce maintenance and reuse ...

With the increasing complexity of hydrologic problems, data collection and data analysis are often carried out in distributed heterogeneous systems. Therefore it is critical for users to determine the origin of data and its trustworthiness. Provenance describes the information life cycle of data products. It has been recognised as one of the most promising methods to improve data transparency. However, due to the complexity of the information life cycle involved, it is a challenge to query the provenance information which may be generated by distributed systems, with different vocabularies and conventions, and may involve knowledge of multiple domains. In this paper, we present a semantic knowledge management framework that tracks and integrates provenance information across distributed heterogeneous systems. It is underpinned by the Integrated Knowledge model that describes the domain knowledge and the provenance information involved in the information life cycle of a particular data product. We evaluate the proposed framework in the context of two real-world water information systems.

This paper presents a brief overview of requirements for development and evaluation of human centred explainable systems. We propose three perspectives on evaluation models for explainable AI that include intrinsic measures, dialogic measures and impact measures. The paper outlines these different perspectives and looks at how the separation might be used for explanation evaluation bench marking and integration into design and development. We propose several avenues for future work.

When designing and implementing real world ambient intelligent systems, we are in need of applicable information systems engineering methods. These should supplement the knowledge engineering tools we can find in the intelligent systems area. The work presented here focuses on explanation-aware ambient intelligent systems. The ability to explain it’s reasoning and actions has been identified as one core capability of any intelligent entity [1]. The question of what is considered a good explanation is context dependent [2], leading to the necessity to design the explanatory capabilities of an ambient intelligent system together with the contextual modelling. We target the requirements elicitation, analysis, and specification processes by making use of a pattern-based approach in form of Jackson’s problem frames [3]. His set of basic problem frames can be extended to be better able to model domain specific aspects. We have previously suggested additional problem frames for explanatory...

C ontext-sensitive processing is crucial in many application domains, not only for mobile and ubiquitous computing, but also for tasks such as collaboration support, intelligent information retrieval, adaptive games, and e-learning. The Modeling and Reasoning in Context (MRC) workshop series, begun in 2004, brings together researchers and practitioners to exchange ideas and results on modeling and reasoning issues for context-sensitive systems. MRC 2007 broadened the focus to also highlight studies of contextualization in human tasks (CHUT), to explore the practical relationships between tasks, actors, and workplace context that may shape system design. The workshop was split into formal paper presentations and discussion sessions. The first two discussions combined themed panels with audience participation, while the closing free-form discussion offered the opportunity for participants to examine issues of their choice and provide closing perspective on the workshop as a whole. Following an MRC tradition, the workshop also included an informal dinner, enabling participants to continue their discussions in a traditional Copenhagen restaurant.

When designing and implementing real world ambient intelligent systems we are in need of applicable information systems engineering methods. The tools we find in the intelligent systems area focus on the knowledge engineering parts, whereas traditional software engineering techniques are usually not designed with the peculiarities of intelligent systems design in mind. This holds in particular for explanation-aware intelligent systems. This work looks at problem frames for explanations and investigates how problem frames can be used to elicit, analyse, and specify these specific requirements. The point of departure is an existing ambient intelligent information system for the hospital ward domain. The work presented here analyses how such a system can be redesigned with a focus on explanation-awareness.

Applications deployed on cyber-infrastructures often rely on multiple data sources and distributed compute resources to access, process, and derive results. When application results are maps, it is possible that non-intentional imperfections can get introduced into the map generation processes because of several reasons including the use of low quality datasets, use of data filtering techniques incompatible for the kind of map to be generated, or even the use of inappropriate mapping parameters, eg, low-resolution gridding ...

Information retrieval and integration systems typically must handle incomplete and inconsistent data. Current approaches attempt to reconcile discrepant information by leveraging data quality, user preferences, or source provenance information. Such approaches may overlook the fact that information is interpreted relative to its context. Therefore, discrepancies may be explained and thereby resolved if contexts are taking into account. In this paper, we describe an information integrator that is capable of explaining ...

Logic-based controlled natural languages usually provide some facility for compositional representation, minimally including sentence level coordination and sometimes subordination. These forms of compositional representation are useful for expressing short passages of a few sentences, but compared to natural language they offer only a limited range of expression, and they are unwieldy for expressing longer passages. This paper describes a method for bringing more naturalness to controlled natural languages. This approach defines a model for representing compositional structures found in natural language, such as Antithesis, Concession, and Unless in a way that is both rhetorically expressive and logically reducible. This is demonstrated through a series of examples using both natural and controlled natural language. The paper then describes a set of intercompositional rules of inference that can be used to reason about the interrelationships between compositional structures, facilitating the discovery and assessment of supportive and conflicting relationships among them. This ability to represent and discover interrelationships among discourse representation structures could be useful for developing applications that must not only follow a set of rules of inference, but reason about the rules themselves.

Abstract. This paper describes a project aiming at enhancing social tagging with reasoning and explanations. So as to keep with the ease of use characteristic of social media, simple explanations are required. A working hypothesis of the work reported in this paper is that simple explanations require simple reasoning. The approach to reasoning presented in this paper is minimalist: First, it precludes involved forms of reasoning such as refutation or excluded middle; second, it does not need (structural) induction. It is furthermore ...

In this paper we describe the application of a novel learning and problem solving architecture to the domain of airspace management, where multiple requests for the use of airspace need to be reconciled and managed automatically. The key feature of our “Generalized Integrated Learning Architecture”(GILA) is a set of integrated learning and reasoning (ILR) systems coordinated by a central meta-reasoning executive (MRE). Each ILR learns independently from the same training example and contributes to problem- ...

Learning Event Models that Explain Anomalies Matthew Molineaux1, David W. Aha2, and Ugur Kuter3 1Knexus Research Corporation; 9120 Beachway Lane; Springfield, VA 22153; USA 2Naval Research Laboratory, Code 5514; Washington DC; USA 3University of Maryland, Institute ...

We consider the problem of automated planning in partiallyobservable dynamic environments, where exogenous events that cannot be directly observed affect the state of the world. In these environments, a planner's knowledge of the world is limited, and state transitions can be both ambiguous and difficult to predict due to that lack of knowledge. We describe a new formalism and new algorithms that enable a planner to proactively expand its knowledge of the environment during planning and execution, by modeling the exogenous events that can occur and forming explanations that reveal information about the world. We have implemented our new algorithms in a variant of the well-known SHOP2 planner that can replan when a failure occurs during plan execution. We have conducted an ablation study in two planning domains to examine the effects of explanation on execution. The results demonstrate that our algorithm successfully increases the performance of an agent using it in two planning domains. This improvement results from the agent having increased knowledge of the environment, which allows it to more accurately predict future events and ultimately make better plans.

Agents with incomplete environment models are likely to be surprised, and this represents an opportunity to learn. We investigate approaches for situated agents to detect surprises, discriminate among different forms of surprise, and hypothesize new models for the unknown events that surprised them. We instantiate these approaches in a new goal reasoning agent (named FOOLMETWICE), investigate its performance in simulation studies, and report that it produces plans with significantly reduced execution cost in comparison to not learning models for surprising events.

Logic-based controlled natural languages usually provide some facility for compositional representation, minimally including sentence level coordination and sometimes subordination. Although these compositional forms suffice for representing short passages, they can become unwieldy for expressing entire paragraphs and documents. This paper describes an approach to representing larger composite texts in a controlled natural language. This approach, called discourse-based reasoning, integrates rhetorical structure theory with argumentation theory to define a model for defining composite structures and argument strategies in an ontological representation. Rhetorical structures are used to represent controlled texts, and argument strategies are defined for reasoning about interactions between structures. This provides the basis for expressing, summarizing, and interacting with explanatory and argumentative discourse. This would expand the scope of problems that may be addressed using controlled natural languages.

Much useful new information (e.g. information in news reports) is often that which is surprising or unexpected. In other words, we harbour many expectations about the world, and when any of these expectations are violated (i.e. made inconsistent) by new information, we have a strong indicator that the information is interesting for us. An expectation can be compared with an integrity constraint. Both an expectation and integrity constraint can be represented by an implicational formula in classical logic, and every time we get new information, we compare it with the implicational formula. However, with an integrity constraint, we are primarily seeking information that is consistent with the implicational formula. In constrast, with an expectation, we are primarily seeking information that is inconsistent with the implicational formula. In this paper, we present a framework for representing and analysing expectations. We consider for an application language the syntax of expectations...

The number of news reports published online is now so great that it is impossible for any person to read all of them. Not all of these reports are equally interesting. Automating the identification and evaluation of interest in news is therefore a potentially valuable goal. Traditional methods of information management such as information filtering, information retrieval and collaborative filtering are unsuited to the task of filtering news. This abstract presents a framework that permits the identification of interesting news by means of the identification of violated expectations. Facts derived from news reports, expectations and related background knowledge can be used to (i) justify the decision to rate news as interesting, (ii) explain why the information in the report is unexpected and, (iii) explain the context in which the report appears.

Applications deployed on cyber-infrastructures often rely on multiple data sources and distributed compute resources to access, process, and derive results. When application results are maps, it is possible that non-intentional imperfections can get introduced into the map generation processes because of several reasons including the use of low quality datasets, use of data filtering techniques incompatible for the kind of map to be generated, or even the use of inappropriate mapping parameters, eg, low-resolution gridding ...

This paper presents a study conducted on the minimum number of open stacks problem (MOSP) which occurs in various production environments where an efficient simultaneous utilization of resources (stacks) is needed to achieve a set of tasks. We investigate through this problem how classical look-back reasonings based on explanations could be used to prune the search space and design a new solving technique. Explanations have often been used to design intelligent backtracking mechanisms in Constraint Programming whereas their use in nogood recording schemes has been less investigated. In this paper, we introduce a generalized nogood (embedding explanation mechanisms) for the MOSP that leads to a new solving technique and can provide explanations. The Minimum number of Open Stacks

With the increasing complexity of hydrologic problems, data collection and data analysis are often carried out in distributed heterogeneous systems. Therefore it is critical for users to determine the origin of data and its trustworthiness. Provenance describes the information life cycle of data products. It has been recognised as one of the most promising methods to improve data transparency. However, due to the complexity of the information life cycle involved, it is a challenge to query the provenance information which may be generated by distributed systems, with different vocabularies and conventions, and may involve knowledge of multiple domains. In this paper, we present a semantic knowledge management framework that tracks and integrates provenance information across distributed heterogeneous systems. It is underpinned by the Integrated Knowledge model that describes the domain knowledge and the provenance information involved in the information life cycle of a particular data product. We evaluate the proposed framework in the context of two real-world water information systems.

Applications deployed on cyber-infrastructures often rely on multiple data sources and distributed compute resources to access, process, and derive results. When application results are maps, it is possible that non-intentional imperfections can get introduced into the map generation processes because of several reasons including the use of low quality datasets, use of data filtering techniques incompatible for the kind of map to be generated, or even the use of inappropriate mapping parameters, eg, low-resolution gridding ...

The information extraction system iDocument interactively extracts information from text such as instances and relations with respect to existing background knowledge. An extraction process creates weighted recommendations describing indications of relevant information. During execution, each process step records its output into an instantiated process model. We reused these bits of information for generating conceptual, functional as well as causal explanations. The purpose of these explanations is to illustrate the evolution of recommendations for convincing users of their validity. In order to visualise explanations, our component utilises different mechanisms for textual, tabular, and graphical rendering styles.

The information extraction system iDocument interactively extracts information from text such as instances and relations with respect to existing background knowledge. An extraction process creates weighted recommendations describing indications of relevant information. During execution, each process step records its output into an instantiated process model. We reused these bits of information for generating conceptual, functional as well as causal explanations. The purpose of these explanations is to illustrate the evolution of recommendations for convincing users of their validity. In order to visualise explanations, our component utilises different mechanisms for textual, tabular, and graphical rendering styles.

The information extraction system iDocument interactively extracts information from texts such as instances and relations with respect to existing background knowledge. An extraction process creates weighted hypotheses describing indications of relevant information. During execution, each process step records its output into an instantiated process model. We reused these bits of information for generating conceptual, functional as well as causal explanations. In order to visualise explanations, our component utilises different mechanisms for textual, explorative, and pictorial rendering styles.

The information extraction system iDocument interactively extracts information from texts such as instances and relations with respect to existing background knowledge. An extraction process creates weighted hypotheses describing indications of relevant information. During execution, each process step records its output into an instantiated process model. We reused these bits of information for generating conceptual, functional as well as causal explanations. In order to visualise explanations, our component utilises different mechanisms for textual, explorative, and pictorial rendering styles.

Communication technology is often used to facilitate the exchange of informa-tion which gives the social networks of a person insight into what the person is doing and what his plans and expectations are. Examples for this are the twitter nanoblog service, www.twitter.com, ...