Domain Ontology learning

We present a novel approach to the automatic acquisition of taxonomies or concept hierarchies from domain-specific texts based on Formal Concept Analysis (FCA). Our approach is based on the assumption that verbs pose more or less strong selectional restrictions on their arguments. The conceptual hierarchy is then built on the basis of the inclusion relations between the extensions of the selectional restrictions of all the verbs, while the verbs themselves provide intensional descriptions for each concept. We formalize this idea in terms of FCA and show how our approach can be used to acquire a concept hierarchy for the tourism domain out of texts. We then evaluate our method by considering an already existing ontology for this domain.

We present a novel approach to the automatic acquisition of taxonomies or concept hierarchies from texts based on Formal Concept Analysis. Our approach is based on the assumption that verbs pose strong selectional restrictions on their arguments. The conceptual hierarchy is then built on the basis of the inclusion relations between the extensions of the selectional restrictions of all the verbs, while the verbs themselves provide intensional descriptions for each concept. We formalize this idea in terms of FCA and show how our approach can be used to acquire a concept hierarchy for the tourism domain out of texts. In particular, we focus on the question if smoothing techniques have an influence on the quality of the generated concept hierarchies. We evaluate our approach by considering an already existing ontology for this domain.

Abstract: We present a novel approach to learning taxonomic relations betweenterms by considering multiple and heterogeneous sources of evidence. In order toderive an optimal combination of these sources, we exploit a machine-learning approach,representing all the sources of evidence as first-order features and trainingstandard classifiers. We consider in particular different features derived from WordNet,an approach matching Hearst-style patterns in a corpus

We present a novel approach to learning taxonomic relations between terms by considering multiple and heterogeneous sources of evidence. In order to derive an optimal combination of these sources, we exploit a machine-learning approach, representing all the sources of evidence as first-order features and training standard classifiers. We consider in particular different features derived from Word-Net, an approach matching Hearst-style patterns in a corpus and on the Web as well as further methods mentioned in the literature. In particular, we explore different classifiers as well as various strategies for dealing with unbalanced datasets. We evaluate our approach by comparing the results with a reference taxonomy for the tourism domain.

As this chapter shows, digital 3D reconstructions of historic architecture serve many purposes in research and related areas. This comprises answering research questions by creating a 3D model, preserving cultural heritage, communicating knowledge in education, and providing a structure for knowledge organization. The process of creating a 3D reconstruction is often challenging, for example, because of lacking or ambiguous sources. In order to create a 3D reconstruction based on scientific values, guidelines, and standards are needed. • How can a 3D reconstruction contribute to knowledge? • Where are 3D reconstructions made? • Why do a 3D reconstruction in a scientific context? • When can the model be reused in other research contexts? • How can one create a scientifically correct 3D reconstruction? • What guidelines are there for preparing a scientific 3D reconstruction?

As this chapter shows, digital 3D reconstructions of historic architecture serve many purposes in research and related areas. This comprises answering research questions by creating a 3D model, preserving cultural heritage, communicating knowledge in education, and providing a structure for knowledge organization. The process of creating a 3D reconstruction is often challenging, for example, because of lacking or ambiguous sources. In order to create a 3D reconstruction based on scientific values, guidelines, and standards are needed. • How can a 3D reconstruction contribute to knowledge? • Where are 3D reconstructions made? • Why do a 3D reconstruction in a scientific context? • When can the model be reused in other research contexts? • How can one create a scientifically correct 3D reconstruction? • What guidelines are there for preparing a scientific 3D reconstruction?

As this chapter shows, digital 3D reconstructions of historic architecture serve many purposes in research and related areas. This comprises answering research questions by creating a 3D model, preserving cultural heritage, communicating knowledge in education, and providing a structure for knowledge organization. The process of creating a 3D reconstruction is often challenging, for example, because of lacking or ambiguous sources. In order to create a 3D reconstruction based on scientific values, guidelines, and standards are needed. • How can a 3D reconstruction contribute to knowledge? • Where are 3D reconstructions made? • Why do a 3D reconstruction in a scientific context? • When can the model be reused in other research contexts? • How can one create a scientifically correct 3D reconstruction? • What guidelines are there for preparing a scientific 3D reconstruction?

The task of binary relation extraction in IE [3] is based mainly on high-frequent verbs and patterns. During the extraction of a specific relation from MEDLINE 1 English abstracts, it is noticed that besides the high-frequent verb itself which represents the specific relation, some other word forms, such as the nominal and adjective forms of this verb, as well as its synonyms, also play a very important role. Because of the characteristics of the sub-language in MEDLINE abstracts, the synonym information of the verb can not be obtained directly from a lexicon such as WordNet 2 [1]. In this paper, an approach which makes use of both corpus information and WordNet synonym set (WN-synset) information is proposed to find out the synonyms of a domain-specific verb in a sub-language. Given a golden standard synonym list obtained from the test corpus, the recall of this approach achieved 60% under the condition that the precision is 100%. The verbs corresponding to the 60% recall cover 93.05% of all occurrences of verbs in the golden standard synonym list.

Abstract-OntoGen is a semi-automatic and data-driven ontology editor focusing on editing of topic ontologies. It utilizes text mining tools to make the ontology-related tasks simpler to the user. This focus on building ontologies from textual data is what we are trying to bridge. We have successfully extended OntoGen to work with image data and allow for ontology construction and editing on collections of labeled or unlabeled images. Browsing large heterogenous image collections efficiently is certainly a challenging task -and we feel that semiautomatic ontology construction, as described in this paper, makes this task easier.

Ontology defines a set of representational primitives with which a domain of knowledge is modeled. Cloud-based computing is an emerging practice that offers significantly more infrastructure and financial flexibility than traditional computing models. Ontologies are playing very important part in many areas such as intelligent information retrieval, knowledge management and organization, electronic commerce. Today, search engine crawlers are retrieving billions of unique URL’s or web page. The main purpose of the Semantic web and ontology is to integrate heterogeneous data and enable interoperability among disparate systems. As the ontology is a standard, explicit and formalized description for shared conceptual model, the educational enterprises can be integrated by applying semantic express, shared knowledge described by ontology and automatic inference mechanism in educational institute management.

The Graphs with Results and Actions Interrelated (GRAI) methodology and its development the GRAI Integrated Methodology (GIM) are established enterprise modelling (EM) techniques for representing the decision architecture of manufacturing systems. However, they lack the support of certain modelling constructs, which in turn can lead to inconsistent and inadequate results. To help address this issue, this paper proposes a classification schema of manufacturing decisions that will facilitate the identification and analysis of decisions for constructing GRAI models. The classification schema is based on: (i) a continuum of organisational decision characteristics; (ii) a categorisation of manufacturing decision domains (DDs) and (iii) a list of manufacturing system configurations. By using this schema during the initial analysis phase and the model construction phase, it is possible to improve the process of identifying key manufacturing decisions and the associated processes, activities and entities.

Today, with the advances of new technologies, accidents, incidents and occupational health records are stored in diverse repositories. The amount of information of HSE that is daily generated has become increasingly huge. Most of this information is stored as unstructured data. The challenge of unstructured data is a top priority for industries that are looking for ways to search, sort, analyze and extract knowledge from masses of data. In this paper we provide with a solution to oil and gas industry for accident investigation using information extraction techniques. The main objective is to propose and evaluate information extraction techniques in occupational health control process, particularly, for automatic detection of accidents from unstructured texts. Our proposal divides the problem in subtasks such as text analysis, recognition and classification of failed occupational health control, resolving accidents.

In dynamic contexts, ontologies and their lexical component (termino-ontologies or TOR) have to frequently adapt to domain evolutions, new uses and new user needs. Among all depending data, ontologybased semantic annotations also are regularly updated to annotate new documents or to reflect new points of view. Within the TextViz ontology-based annotation framework, we propose the EvOnto tool and method that supports a coherent joint change management of termino-ontologies and semantic annotations as well as quality criteria to evaluate automatic text annotations and to detect lacks in the ontology.

In this paper we propose a system for the recommendation of tagged pictures obtained from the Web. The system, driven by user feedback, executes an abductive reasoning (based on WordNet synset semantic relations) that is able to iteratively lead to new concepts which progressively represent the cognitive creative user state. Furthermore we design a selection mechanism to pick the most relevant abductive inferences by mixing a topological graph analysis together with a semantic similitude measure.

A new line of investigation that integrates studies on artificial intelligence and Internet technolo gies, which is known as the Semantic Web, is presented. A review of the present state of research is given; problems on the establishment of knowledge spaces on the Internet, the means and methods for the extrac tion of knowledge from texts in natural languages, as well as questions on the use of knowledge spaces in the creation of applied intelligent systems operating on the Internet, are considered.

Since the 1990s, the application of digital 3D reconstruction and computer-based visualization of cultural heritage has increased. Virtual reconstruction and 3D visualization have revealed a new "glittering" research space for object-oriented disciplines such as archaeology, art history and architecture. Nevertheless, the scientists concerned with the new technology soon recognized the lack of documentation standards in the 3D projects, leading to the loss of information, findings and the fusion of knowledge behind the digital 3D representation. Based on the methodological fundamentals of digital 3D reconstruction, the potentials and challenges in the light of emerging Semantic Web and Web3D technologies is introduced here. The presentation describes a scientific methodology and a collaborative web-based research environment followed by crucial features for these kinds of projects. As the groundwork, a human-and machine-readable "language of objects" and the implementation of these semantic patterns for spatial research purposes on destroyed and/or never realized tangible cultural heritage will be discussed. Using examples from the practice, the presentation explains the requirements of the Semantic Web (Linked Data), the role of controlled vocabularies, the architecture of the VRE and the impact of a customized integration of interactive 3D models within WebGL technology. The presentation intends to showcase the state-of-the-art steps on the way to a digital research infrastructure. The focus lies on the introduction of scholarly approved and sustainable digital 3D reconstruction, compliant with recognized documentation standards and following the Linked Data requirements.

Ontologies are being used to organize information in many domains like artificial intelligence, information science, semantic web, library science. Ontologies of an entity having different information can be merged to create more knowledge of that particular entity. Ontologies today are powering more accurate search and retrieval in websites like Wikipedia etc. As we move towards the future to Web 3.0, also termed as the semantic web, ontologies will play a more important role. Ontologies are represented in various forms like RDF, RDFS, XML, OWL etc. Querying ontologies can yield basic information about an entity. This paper proposes an automated method for ontology creation, using concepts from NLP (Natural Language Processing), Information Retrieval and Machine Learning. Concepts drawn from these domains help in designing more accurate ontologies represented using the XML format. This paper uses document classification using classification algorithms for assigning labels to documents, document similarity to cluster similar documents to the input document, together, and summarization to shorten the text and keep important terms essential in making the ontology. The module is constructed using the Python programming language and NLTK (Natural Language Toolkit). The ontologies created in XML will convey to a lay person the definition of the important term's and their lexical relationships.

Documents contain textual information, which is of the utmost importance for all the organizations. Document management systems have been used to store vast amounts of unstructured textual data described with minimal metadata, a method that has several limitations. In order to convert hidden knowledge into machine-readable data with rich connections, this paper presents work in progress on the development of the first end-to-end guided approach to construct a Knowledge Graph from Greek government documents from the Greek open government portal. The resulted Knowledge graph consists a proof-of-concept graph, that illustrates the beneficial semantic relationships between the textual data.

This paper presents an experimental evaluation of five ontology construction tools that we are using for HIV/AIDS FAQ retrieval system. Ontologies have been widely used in natural language processing applications especially in question answering systems. As the construction of ontologies is a time consuming and error prone task, the use of tools would be very helpful. To compare the five tools a knowledge source on HIV/AIDS was collected from MASA Antiretroviral Therapy booklet which consists of 205 most frequently asked questions About HIV/AIDS and ARV Therapy. The booklet was prepared by the Ministry of Health in Botswana. The quality of the resulting ontology along with other features of the tools was evaluated. The attributes that we specifically looked at for each tool include file formats supported, concepts extraction techniques used, ontology visualization support, and ontology export formats. Based on the experiments we carried out, Text2Onto, which is the latest version of Text-To-Onto, is found to be most appropriate and will be used to construct the HIV/AIDS ontology for FAQ retrieval. Text2Onto, which integrates Text-To-ONTO and GATE provides ranking of terms, detecting changes in knowledge source, automatic update of ontology, and tracking modifications to ontology.

In this paper, the population ontology problem is addressed and a semiautomatic methodology for ontology learning is proposed. In this work a study of advances in this research area is presented, which is divided according with the aim of the ontology and the class extraction, creation, or population. Based on this study, an initial proposal for ontology semiautomatic population is introduced. This initial proposal consists of four general steps: class extraction, creation, population and evaluation, where the population step is the main objective of this research. As future work, this methodology is intended to be applied to pedagogic domain, specially to classroom learning tools.

Due to the explosive growth of the Web, the domain of Web personalization has gained great momentum both in the research and commercial areas. One of the most popular web personalization systems is recommender systems. In recommender systems choosing user information that can be used to profile users is very crucial for user profiling. In Web 2.0, one facility that can help users organize Web resources of their interest is user tagging systems. Exploring user tagging behavior provides a promising way for understanding users' information needs since tags are given directly by users. However, free and relatively uncontrolled vocabulary makes the user self-defined tags lack of standardization and semantic ambiguity. Also, the relationships among tags need to be explored since there are rich relationships among tags which could provide valuable information for us to better understand users. In this paper, we propose a novel approach for learning tag ontology based on the widely used lexical database WordNet for capturing the semantics and the structural relationships of tags. We present personalization strategies to disambiguate the semantics of tags by combining the opinion of WordNet lexicographers and users' tagging behavior together. To personalize further, clustering of users is performed to generate a more accurate ontology for a particular group of users. In order to evaluate the usefulness of the tag ontology, we use the tag ontology in a pilot tag recommendation experiment for improving the recommendation performance by exploiting the semantic information in the tag ontology. The initial result shows that the personalized information has improved the accuracy of the tag recommendation.

Recently, user tagging systems have grown in popularity on the web. The tagging process is quite simple for ordinary users, which contributes to its popularity. However, free vocabulary has lack of standardization and semantic ambiguity. It is possible to capture the semantics from user tagging into some form of ontology, but the application of the resulted ontology for recommendation making has not been that flourishing. In this paper we discuss our approach to learn domain ontology from user tagging information and apply the extracted tag ontology in a pilot tag recommendation experiment. The initial result shows that by using the tag ontology to re-rank the recommended tags, the accuracy of the tag recommendation can be improved.

The Graphs with Results and Actions Interrelated (GRAI) methodology and its development the GRAI Integrated Methodology (GIM) are established enterprise modelling (EM) techniques for representing the decision architecture of manufacturing systems. However, they lack the support of certain modelling constructs, which in turn can lead to inconsistent and inadequate results. To help address this issue, this paper proposes a classification schema of manufacturing decisions that will facilitate the identification and analysis of decisions for constructing GRAI models. The classification schema is based on: (i) a continuum of organisational decision characteristics; (ii) a categorisation of manufacturing decision domains (DDs) and (iii) a list of manufacturing system configurations. By using this schema during the initial analysis phase and the model construction phase, it is possible to improve the process of identifying key manufacturing decisions and the associated processes, activities and entities.

Comprehensive, consistent and standardized definition of terms in a specific area are crucial for ensuring interoperability among IT applications related to this area. This goal can be achieved in developing central terminological resources in the form of thesauruses or ontologies, which can then be used for deriving terminology in specific sub-fields. In this paper we describe how RudOnto, a terminological resource for mining engineering developed at the University of Belgrade Faculty of Mining and Geology, can be used for generating a mining equipment ontology. RudOnto is an integral part of a specialized mining engineering information system, which features functionalities for management and maintaining of RudOnto, as well as a wizard for exporting subsets of its concepts to several specific formats, including the web ontology language OWL.

By amassing 'wisdom of the crowd', social tagging systems draw more and more academic attention in interpreting Internet folk knowledge. In order to uncover their hidden semantics, several researches have attempted to induce an ontology-like taxonomy from tags. As far as we know, these methods all need to compute an overall or relative generality for each tag, which is difficult and error-prone. In this paper, we propose an agglomerative hierarchical clustering framework which relies only on how similar every two tags are. We enhance our framework by integrating it with a topic model to capture thematic correlations among tags. By experimenting on a designated online tagging system, we show that our method can disclose new semantic structures that supplement the output of previous approaches. Finally, we demonstrate the effectiveness of our method with quantitative evaluations.

Modern information systems are changing the idea of "data processing" to the idea of "concept processing", meaning that instead of processing words, such systems process semantic concepts which carry meaning and share contexts with other concepts. Ontology is commonly used as a structure that captures the knowledge about a certain area via providing concepts and relations between them. Traditionally, concept hierarchies have been built manually by knowledge engineers or domain experts. However, the manual construction of a concept hierarchy suffers from several limitations such as its coverage and the enormous costs of its extension and maintenance. Ontology learning, usually referred to the (semi-)automatic support in ontology development, is usually divided into steps, going from concepts identification, passing through hierarchy and non-hierarchy relations detection and, seldom, axiom extraction. It is reasonable to say that among these steps the current front...

Ontologies are an important part of the Semantic Web as well as of many intelligent systems. However, the traditional expert-driven development of ontologies is time-consuming and often results in incomplete and inappropriate ontologies. In addition, since ontology evolution is not controlled by end users, it may take too long for a conceptual change in the domain to be reflected in the ontology. In this paper, we present a recommendation algorithm in a Web 2.0 platform that supports end users to collaboratively evolve ontologies by suggesting semantic relations between new and existing concepts. We use the Wikipedia category hierarchy to evaluate our algorithm and our experimental results show that the proposed algorithm produces high quality recommendations.

Ontologies are used by modern knowledge-based systems to represent and share knowledge about an application domain. Ontology population looks for identifying instances of concepts and relationships of an ontology. Manual population by domain experts and knowledge engineers is an expensive and time consuming task so, automatic and semi-automatic approaches are needed. This article proposes an initial approach for automatic ontology population from textual sources that use natural language processing and machine learning techniques. Some experiments using a family law corpus were conducted in order to evaluate it. Initial results are promising and indicate that our approach can extract instances with good effectiveness. 2 AN ONTOLOGY DEFINITION Ontologies are formal specifications of concepts in a domain of interest. Their classes, relationships,

This article presents the combination of Latent Semantic Analysis (LSA) with other natural language processing techniques (stemming, removal of closed-class words and word sense disambiguation) to improve the automatic assessment of students' free-text answers. The combinational schema has been tested in the experimental framework provided by the free-text Computer Assisted Assessment (CAA) system called Atenea (Alfonseca & Pérez, 2004). This system is able to ask randomly or according to the students' profile an openended question to the student and then, assign a score to it. The results prove that for all datasets, when the NLP techniques are combined with LSA, the Pearson correlation between the scores given by Atenea and the scores given by the teachers for the same dataset of questions improves. We believe that this is due to the complementarity between LSA, which works more at a shallow semantic level, and the rest of the NLP techniques used in Atenea, which are more focused on the lexical and syntactical levels.

Ontologies are used by modern knowledge-based systems to represent and share knowledge about an application domain. Ontology population looks for identifying instances of concepts and relationships of an ontology. Manual population by domain experts and knowledge engineers is an expensive and time consuming task so, automatic and semi-automatic approaches are needed. This article proposes an initial approach for automatic ontology population from textual sources that use natural language processing and machine learning techniques. Some experiments using a family law corpus were conducted in order to evaluate it. Initial results are promising and indicate that our approach can extract instances with good effectiveness. 2 AN ONTOLOGY DEFINITION Ontologies are formal specifications of concepts in a domain of interest. Their classes, relationships,

We propose a method for semi-automatic construction of an ontology of a given branch of science for measuring its evolution in time. The method relies on a collection of documents in the given thematic domain. We observe that the words of different levels of abstraction are located within different parts of a document: say, the title or abstract contains more general words than the body of the paper. What is more, the hierarchical structure of the documents allows us to determine the parent-child relation between words: e.g., a word that appears in the title of a paper is a candidate for a parent of the words appearing in the body of this paper; if such a relation is repeated several times, we register such a parent-child pair in our ontology. Using the papers corresponding to different years, we construct such an ontology for each year independently. Comparing such ontologies (using tree edit distance measure) for different years reveals the trends of evolution of the given branch of science.

Automated Essay Grading (AEG) is a very important research area in educational technology. Latent Semantic Analysis (LSA) is an Information Retrieval (IR) technique used for automated essay grading. LSA forms a word by document matrix and then the matrix is decomposed using Singular Value Decomposition (SVD) technique. Existing AEG systems based on LSA cannot achieve higher level of performance to be a replica of human grader. We have developed an AEG system using Generalized Latent Semantic Analysis (GLSA) which makes n-gram by document matrix instead of word by document matrix. The system has been evaluated using details representation. Experimental results show that the proposed AEG system achieves higher level of accuracy as compared to human grader.

Collaborative tagging systems have recently emerged as a powerful way to label and organize large collections of data. The informal social classification structure in these systems, also known as folksonomy, provides a convenient way to annotate resources by allowing users to use any keyword or tag that they find relevant. Although folksonomies and the respective tags often lack a context-independent and intersubjective definition of meaning, the assumption that the evolving structure of these digital records contains implicit evidences for the underlying semantics has been proven by successful approaches of making the emergent semantics explicit. In this paper we propose an approach for extracting ontological structures from folksonomies that exploits the power of fuzzy clustering using new similarity and generality measure. The fuzzy clustering process discovers ambiguous tags and disambiguates them all at once, and the new similarity measure gives more accurate results as it calculates co-occurrences based on distinct users and not only in the number of co-occurrences of two distinct words. The generated ontology can be used to enhance various tasks in the tagging systems, such as tag disambiguation, result visualization, and ontology evolution. Our experimental results on real world data sets show that our method can effectively learn the ontology structure from the folksonomies.

To tackle the issues of semantic collision and inconsistencies between ontologies and the original data model while learning ontology from relational database (RDB), a semi-automatic semantic consistency checking method based on graph intermediate representation and model checking is presented. Initially, the W-Graph, as an intermediate model between databases and ontologies, was utilized to formalize the semantic correspondences between databases and ontologies, which were then transformed into the Kripke structure and eventually encoded with the SMV program. Meanwhile, description logics (DLs) were employed to formalize the semantic specifications of the learned ontologies, since the OWL DL showed good semantic compatibility and the DLs presented an excellent expressivity. Thereafter, the specifications were converted into a computer tree logic (CTL) formula to improve machine readability. Furthermore, the task of checking semantic consistency could be converted into a global mode...

El problema que se detectó en la Especialidad de Matemática y Física de la UNHEVAL es que los alumnos con mucha dificultad podían determinar el dominio y rango de las funciones de manera analítica; ello constituye un problema de magnitud, porque serán docentes de matemática y física; o sea, docentes sin capacidades específicas ejerciendo la docencia para estudiantes de Educación Básica en plena formación, en la cultura matemática; por tanto, es probable que esto sea uno de las factores de la deficiencia en el aprendizaje de los alumnos de educación básica. Debido a lo dicho y tratando de paliar en algo el problema de aprendizaje del dominio y rango de funciones, en la investigación, se propone la aplicación del método gráfico con la ayuda del programa Geogebra, entonces en función a ello, el informe final ha sido diseñado de la siguiente manera: Capítulo I, incluye todo lo referente al problema de investigación como, descripción del problema, formulación del problema de investigación, objetivos, hipótesis, justificación e importancia, viabilidad, limitaciones, etc. Capítulo II, incluye el marco teórico, donde está considerado: los antecedentes de la investigación, las teorías básicas y la definición conceptual de términos usados en la investigación. Capítulo III, en este bloque está considerado todo lo referente al marco metodológico de la investigación, que son el tipo de investigación, diseño y esquema, población y muestra, instrumentos de recolección de datos, y las técnicas para el análisis y procesamiento de los datos. Capítulo IV, se considera los resultados obtenidos en el trabajo de campo, debidamente procesados con un analizador estadístico; en esta parte se presenta la aplicación de la estadística descriptiva y la estadística inferencial con la prueba de hipótesis para la diferencia de medias, dicho estadígrafo permitió el contraste del objetivo general o la hipótesis general. Luego, está incluido la discusión de resultados donde se analiza lo hallado durante el trabajo de campo, y en lo posible, está contrastado con referencias bibliográficas; además, están las conclusiones, sugerencias, la bibliografía y los anexos.

Knowledge engineers have had difficulty in automatically constructing and populating domain ontologies, mainly due to the well-known knowledge acquisition bottleneck. In this paper, we attempt to alleviate this problem by proposing an iterative unsupervised approach to identifying and extracting ontological class instances from the Web. The proposed approach considers the Web as a big corpus and relies on a confidence-weighted metric based on semantic measures and web-scale statistics as types of evidence. Moreover, our iterative method is able to learn, to some extent, domain-specific linguistic patterns for extracting ontological class instances. We obtained encouraging results for the final ranking of candidate instances as well as an accuracy performance up to 97% for the patterns found by our method.

Evaluating a taxonomy learned automatically against an existing gold standard is a very complex problem, because differences stem from the number, label, depth and ordering of the taxonomy nodes. In this paper we propose casting the problem as one of comparing two hierarchical clusters. To this end we defined a variation of the Fowlkes and Mallows measure (Fowlkes and Mallows, 1983). Our method assigns a similarity value Bi(l, r) to the learned (l) and reference (r) taxonomy for each cut i of the corresponding anonymised ...

This paper proposes a surrounding word sense model (SWSM) that uses the distribution of word senses that appear near ambiguous words for unsupervised all-words word sense disambiguation in Japanese. Although it was inspired by the topic model, ambiguous Japanese words tend to have similar topics since coarse semantic polysemy is less likely to occur than that in Western languages as Japanese uses Chinese characters, which are ideograms. We thus propose a model that uses the distribution of word senses that appear near ambiguous words: SWSM. We embedded the concept dictionary of an Electronic Dictionary Research (EDR) electronic dictionary in the system and used the Japanese Corpus of EDR for the experiments, which demonstrated that SWSM outperformed a system with a random baseline and a system that used a topic model called Dirichlet Allocation with WORDNET (LDAWN), especially when there were high levels of entropy for the word sense distribution of ambiguous words.

Automated Essay Grading (AEG) is a very important research area in educational technology. Latent Semantic Analysis (LSA) is an Information Retrieval (IR) technique used for automated essay grading. LSA forms a word by document matrix and then the matrix is decomposed using Singular Value Decomposition (SVD) technique. Existing AEG systems based on LSA cannot achieve higher level of performance to be a replica of human grader. We have developed an AEG system using Generalized Latent Semantic Analysis (GLSA) which makes n-gram by document matrix instead of word by document matrix. The system has been evaluated using details representation. Experimental results show that the proposed AEG system achieves higher level of accuracy as compared to human grader.

In recommender systems based on multidimensional data, additional metadata provides algorithms with more information for better understanding the interaction between users and items. However, most of the profiling approaches in neighbourhood-based recommendation approaches for multidimensional data merely split or project the dimensional data and lack the consideration of latent interaction between the dimensions of the data. In this paper, we propose a novel user/item profiling approach for Collaborative Filtering (CF) item recommendation on multidimensional data. We further present incremental profiling method for updating the profiles. For item recommendation, we seek to delve into different types of relations in data to understand the interaction between users and items more fully, and propose three multidimensional CF recommendation approaches for top-N item recommendations based on the proposed user/item profiles. The proposed multidimensional CF approaches are capable of incorporating not only localized relations of user-user and/or item-item neighbourhoods but also latent interaction between all dimensions of the data. Experimental results show significant improvements in terms of recommendation accuracy .

The growing popularity of social tagging systems promises to alleviate the knowledge bottleneck that slows down the full materialization of the Semantic Web since these systems allow ordinary users to create and share knowledge in a simple, cheap, and scalable representation, usually known as folksonomy. However, for the sake of knowledge workflow, one needs to find a compromise between the uncontrolled nature of folksonomies and the controlled and more systematic vocabulary of domain experts. In this paper we propose to address this concern by devising a method that automatically enriches a folksonomy with domain expert knowledge and by introducing a novel algorithm based on frequent itemset mining techniques to efficiently learn an ontology over the enriched folksonomy. In order to quantitatively assess our method, we propose a new benchmark for task-based ontology evaluation where the quality of the ontologies is measured based on how helpful they are for the task of personalized information finding. We conduct experiments on real data and empirically show the effectiveness of our approach.

The current way of checking subjective paper is adverse. Evaluating the Subjective Answers is a critical task to perform. When human being evaluates anything, the quality of evaluation may vary along with the emotions of Person. In Machine Learning, all result is only based on the input data provided by the user. Our proposed system uses machine learning and NLP to solve this problem. Our Algorithm performs a task like Tokenizing words and sentences, Part of Speech tagging, Chunking, Chinking, Lemmatizing words and Wordnetting to evaluate the subjective answer. Along with it, our proposed algorithm provides the semantic meaning of the context. Our System is divided into two modules. The first one is extracting the data from the scanned images and organizing it in the proper manner and the second is applying ML and NLP to the text retrieved from the above step and giving marks to them.

This paper presents a browser-based semi-automatic taxonomy construction tool Vd-chuck which is able to incorporate text and data mining algorithms into a userfriendly interface. The presented system is browserbased. Its unsupervised learning for concept suggestion and different visualization techniques assist the user with textual and numerical data analysis. We tested the Vdchuck system on a real-world domain: a corpus of documents taken from Slovenian Language technologies conferences. The results show that with our system similar taxonomies as with other taxonomy editors can be constructed.

For decision makers and researchers working in a technical domain, understanding the state of their area of interest is of the highest importance. For this reason, we consider in this chapter, a novel framework for Web-based technology forecasting using bibliometrics (i.e. the analysis of information from trends and patterns of scientific publications). The proposed framework consists of a few conceptual stages based on a data acquisition process from bibliographic online repositories: extraction of domain-relevant keywords, the generation of taxonomy of the research field of interests and the development of early growth indicators which helps to find interesting technologies in their first phase of development. To provide a concrete application domain for developing and testing our tools, we conducted a case study in the field of renewable energy and in particular one of its subfields: Waste-to-Energy (W2E). The results on this particular research domain confirm the benefit of our ...

Knowledge engineers have had difficulty in automatically constructing and populating domain ontologies, mainly due to the well-known knowledge acquisition bottleneck. In this paper, we attempt to alleviate this problem by proposing an iterative unsupervised approach to identifying and extracting ontological class instances from the Web. The proposed approach considers the Web as a big corpus and relies on a confidence-weighted metric based on semantic measures and web-scale statistics as types of evidence. Moreover, our iterative method is able to learn, to some extent, domain-specific linguistic patterns for extracting ontological class instances. We obtained encouraging results for the final ranking of candidate instances as well as an accuracy performance up to 97% for the patterns found by our method.

Users organize personal information in various ways. We believe that this process could be expedited and improved by using domain ontologies. The main problem with this idea is the lack of automatic tools that help nonexpert users to build and maintain their own ontologies. In this study we report progress in the process of adapting ontologies to better represent a given set of documents centered on a topic of interest. More specifically we investigate automatic approaches to enhance the representation of the concepts within the domain ontology. We show that our approach can enrich the vocabulary of each concept with words mined from the set of small documents provided. The method we propose is based on efficient text mining approaches combined with semantic information from WordNet.

Semiotics is a field where research on Computer Science methodologies has focused, mainly concerning Syntax and Semantics. These methodologies, however, are lacking of some flexibility for the continuously evolving web community, in which the knowledge is classified with tags rather than with ontologies. In this paper we propose a multi-agent system for the recommendation of tagged pictures obtained from mainstream Web applications. The agents in this system execute a hybrid reasoning based on WordNet ...

Abstract—Automatic keyphrase extraction is a useful tool in many text related applications, such as clustering and summarisation. Given such importance, systems capable of automatically extracting and representing keyphrases play an important role in Natural Language Processing. In this paper, we present a system with the purpose of extracting keyphrases from heterogeneous Web sources and formally represent them. After exploiting a popular algorithm (KEA), the outcomes revealed to be promising. These keyphrases are ...

In this paper, we propose an automatic and unsupervised methodology to obtain taxonomies of terms from the Web and represent retrieved web sites into a meaningful organization for a desired domain without previous knowledge. It is based on the intensive use of web search engines to retrieve domain suitable resources from which extract knowledge, and to obtain web scale statistics from which infer knowledge relevancy. Results can be useful for easing the access to the web resources or as the first step for constructing ontologies suitable for the Semantic Web.