In order to do more semantics-based information extraction, we require specialized domain models. We develop a hybrid approach for constructing such a domain-specific ontology, which integrates key concepts from the protein-protein–interaction domain with the Gene Ontology. In addition, we present a method for using the domain-specific ontology in a discourse-based analysis module for analyzing full-text articles on protein interactions. The analysis module uses a lexical chaining technique to extract strings of semantically related words that represent the topic structure of the text. We show that the domain-specific ontology improved the performance of the lexical-chaining module. As well the topic structure as represented by the lexical chains contains important information on protein-protein interactions appearing in the same textual context.

In order to do more semantics-based information extraction, we require specialized domain models. We develop a hybrid approach for constructing such a domain-specific ontology, which integrates key concepts from the protein-protein–interaction domain with the Gene Ontology. In addition, we present a method for using the domain-specific ontology in a discourse-based analysis module for analyzing full-text articles on protein interactions. The analysis module uses a lexical chaining technique to extract strings of semantically related words that represent the topic structure of the text. We show that the domain-specific ontology improved the performance of the lexical-chaining module. As well the topic structure as represented by the lexical chains contains important information on protein-protein interactions appearing in the same textual context.

<p>With the growth of information stored over Internet, especially in the biological field, and with discoveries being made daily in this domain, scientists are faced with an overwhelming amount of articles. Reading all published articles is a tedious and time-consuming process. Therefore a way to summarise the information in the articles is needed. A solution is the derivation of an ontology representing the knowledge enclosed in the set of articles and allowing to browse through them. </p><p>In this thesis we present the tool Ontolo, which allows to build an initial ontology of a domain by inserting a set of articles related to that domain in the system. The quality of the ontology construction has been tested by comparing our ontology results for keywords to the ones provided by the Gene Ontology for the same keywords. </p><p>The obtained results are quite promising for a first prototype of the system as it finds many common terms on both ontologies for justa few hundred of inserted articles.</p>

<p>Though a constantly increasing number of ontologies are now available on the Internet, the ontology construction process remains generally a manual task, so consequently an effort demanding task. As no unified ontology construction method is available in the literature, researchers started investigating different frameworks for automatically generating ontologies and, therefore shorten the time required for their construction. This master’s thesis presents a prototype system for automatic construction of ontology, based on ontology design patterns and unstructured texts, such as natural language texts. The use of ontology design patterns allow constructing well structured ontologies and reducing the demand of knowledge experts. A difference between our prototype system and the systems presented in the literature is, the possibility to increase accuracy of the generated ontology by selecting the more relevant terms and associations from the unstructured text and match them against the ontology design patterns. Also, a matching score is introduced to define the level of similarity between the terms extracted and the ontology design patterns. By setting a threshold value on the matching score, the relevant ontology design patterns are selected and used for the ontology construction process. The ontology construction framework used by the prototype system has been developed by the research group in Information Engineering of the School of Engineering, Jönköping University.</p>

<p>Though a constantly increasing number of ontologies are now available on the Internet, the ontology construction process remains generally a manual task, so consequently an effort demanding task. As no unified ontology construction method is available in the literature, researchers started investigating different frameworks for automatically generating ontologies and, therefore shorten the time required for their construction. This master’s thesis presents a prototype system for automatic construction of ontology, based on ontology design patterns and unstructured texts, such as natural language texts. The use of ontology design patterns allow constructing well structured ontologies and reducing the demand of knowledge experts. A difference between our prototype system and the systems presented in the literature is, the possibility to increase accuracy of the generated ontology by selecting the more relevant terms and associations from the unstructured text and match them against the ontology design patterns. Also, a matching score is introduced to define the level of similarity between the terms extracted and the ontology design patterns. By setting a threshold value on the matching score, the relevant ontology design patterns are selected and used for the ontology construction process. The ontology construction framework used by the prototype system has been developed by the research group in Information Engineering of the School of Engineering, Jönköping University.</p> / <p>Även fast ett ökande antal ontologier är tillgängliga på Internet, är ontologikonstruktion fortfarande till största delen en manuell process, som därigenom kräver en stor arbetsinsats. Eftersom ingen enhetlig metodologi för att konstruera ontologier finns i litteraturen, forskare började undersöka olika ansatser för att automatiskt generera ontologier och därigenom förkorta konstruktionstiden. Detta examensarbete presenterar ett prototypsystem för automatisk konstruktion av ontologier, baserat på designmönster för ontologier och ostrukturerad text (text i naturligt språk). Att använda designmönster ger en välstrukturerad ontologi och minskar behovet av expertkunskap. En skillnad mellan vårt system och system i litteraturen är möjligheten att få en mer korrekt ontologi genom att välja de mestrelevanta termerna och relationerna från texterna och matcha dem motdesignmönstren. Ett värde för överensstämmelsen har införts för att kunna beskriva hur stor likhet som finns mellan termerna och designmönstren. Genom att sätta ett tröskelvärde väljs de relevanta designmönstren ut och används för att konstruera ontologin. Den generella processen för ontologikonstruktion som används av prototypsystemet har utvecklats av forskningsgruppen i Informationsteknik vid Ingenjörshögskolan i Jönköping.</p>

Submitted by Antonio Marin Neto (netomarin@gmail.com) on 2018-09-06T00:01:18Z No. of bitstreams: 1 Ontologias na representacao do conhecimeno.pdf: 1512033 bytes, checksum: 5cb40c53baa28e682303bb84ed68e480 (MD5) / Approved for entry into archive by Satie Tagara (satie@marilia.unesp.br) on 2018-09-06T16:36:29Z (GMT) No. of bitstreams: 1 marinneto_a_me_mar.pdf: 1512033 bytes, checksum: 5cb40c53baa28e682303bb84ed68e480 (MD5) / Made available in DSpace on 2018-09-06T16:36:29Z (GMT). No. of bitstreams: 1 marinneto_a_me_mar.pdf: 1512033 bytes, checksum: 5cb40c53baa28e682303bb84ed68e480 (MD5) Previous issue date: 2018-08-06 / Não recebi financiamento / As ontologias surgiram a partir de discussões na Filosofia com o objetivo de discutir a essência das coisas, nos estudos da metafísica por Aristóteles, que dentre os diferentes ramos de investigação filosófica, tratava do conhecimento da essência de toda a realidade. A ontologia trata do estudo do “Ser” enquanto ser. O termo foi posteriormente utilizado em diferentes áreas como a Ciência da Computação (CC), Psicologia, Ciência da Informação (CI), entre outros. Nesse texto é apresentado um levantamento da utilização das ontologias na representação do conhecimento, identificando e destacando a forma de utilização da ontologia como ferramenta na CC e CI. Para tal, o trabalho inicial com uma contextualização do termo ontologia, desde a sua origem na Filosofia até a sua utilização em algumas das áreas do conhecimento. Em seguida é apresentado um levantamento sobre a representação do conhecimento, a destacar a CC e CI, bem como sobre a utilização da ontologia nesse contexto informacional. Afim de evidenciar a importância das ontologias para representação do conhecimento, também são listadas as principais metodologias de criação de ontologia, aprendizado de ontologia, bem como as ferramentas utilizadas para essas tarefas, além do gerenciamento e engenharia de ontologias. E por fim são apresentadas as considerações finais acerca da utilização das ontologias como ferramentas semânticas na CI e como as futuruas pesquisas podem se beneficiar desse trabalho como uma base para o entendimento das ontologias como ferramenta semântica na CI. / Ontologies come from discussions in Philosophy to debate the essence of things in Aristotle's studies of metaphysics, which among the different branches of philosophical inquiry dealt with the knowledge of the essence of all reality. Ontology deals with the study of "Being" as being. The term was later used in different areas such as Computer Science (CC), Psychology, Information Science (IS), andothers. This thesis presents a survey of the use of ontologies in knowledge representation, identifying and highlighting how ontology is used as a semantic tool in CC and IS. For this, is presented a contextualization of the term ontology, from its origin in Philosophy to its use in some of the areas of knowledge. Next, a survey is presented on the representation of knowledge, highlighting the CC and IS, as well as on the use of the ontology in this informational context. In order to demonstrate the importance of ontologies for knowledge representation, the main methodologies for ontology creation, ontology learning, as well as the tools used for these tasks, as well as the management and engineering of ontologies are also listed. Finally, it's present the final considerations about the use of ontologies as semantic tools in IS and how future researches can benefit from this work as a basis for the understanding of ontologies as a semantic tool in IS.

Though a constantly increasing number of ontologies are now available on the Internet, the ontology construction process remains generally a manual task, so consequently an effort demanding task. As no unified ontology construction method is available in the literature, researchers started investigating different frameworks for automatically generating ontologies and, therefore shorten the time required for their construction. This master’s thesis presents a prototype system for automatic construction of ontology, based on ontology design patterns and unstructured texts, such as natural language texts. The use of ontology design patterns allow constructing well structured ontologies and reducing the demand of knowledge experts. A difference between our prototype system and the systems presented in the literature is, the possibility to increase accuracy of the generated ontology by selecting the more relevant terms and associations from the unstructured text and match them against the ontology design patterns. Also, a matching score is introduced to define the level of similarity between the terms extracted and the ontology design patterns. By setting a threshold value on the matching score, the relevant ontology design patterns are selected and used for the ontology construction process. The ontology construction framework used by the prototype system has been developed by the research group in Information Engineering of the School of Engineering, Jönköping University. / Även fast ett ökande antal ontologier är tillgängliga på Internet, är ontologikonstruktion fortfarande till största delen en manuell process, som därigenom kräver en stor arbetsinsats. Eftersom ingen enhetlig metodologi för att konstruera ontologier finns i litteraturen, forskare började undersöka olika ansatser för att automatiskt generera ontologier och därigenom förkorta konstruktionstiden. Detta examensarbete presenterar ett prototypsystem för automatisk konstruktion av ontologier, baserat på designmönster för ontologier och ostrukturerad text (text i naturligt språk). Att använda designmönster ger en välstrukturerad ontologi och minskar behovet av expertkunskap. En skillnad mellan vårt system och system i litteraturen är möjligheten att få en mer korrekt ontologi genom att välja de mestrelevanta termerna och relationerna från texterna och matcha dem motdesignmönstren. Ett värde för överensstämmelsen har införts för att kunna beskriva hur stor likhet som finns mellan termerna och designmönstren. Genom att sätta ett tröskelvärde väljs de relevanta designmönstren ut och används för att konstruera ontologin. Den generella processen för ontologikonstruktion som används av prototypsystemet har utvecklats av forskningsgruppen i Informationsteknik vid Ingenjörshögskolan i Jönköping.

With the growth of information stored over Internet, especially in the biological field, and with discoveries being made daily in this domain, scientists are faced with an overwhelming amount of articles. Reading all published articles is a tedious and time-consuming process. Therefore a way to summarise the information in the articles is needed. A solution is the derivation of an ontology representing the knowledge enclosed in the set of articles and allowing to browse through them. In this thesis we present the tool Ontolo, which allows to build an initial ontology of a domain by inserting a set of articles related to that domain in the system. The quality of the ontology construction has been tested by comparing our ontology results for keywords to the ones provided by the Gene Ontology for the same keywords. The obtained results are quite promising for a first prototype of the system as it finds many common terms on both ontologies for justa few hundred of inserted articles.

Though a constantly increasing number of ontologies are now available on the Internet, the ontology construction process remains generally a manual task, so consequently an effort demanding task. As no unified ontology construction method is available in the literature, researchers started investigating different frameworks for automatically generating ontologies and, therefore shorten the time required for their construction. This master’s thesis presents a prototype system for automatic construction of ontology, based on ontology design patterns and unstructured texts, such as natural language texts. The use of ontology design patterns allow constructing well structured ontologies and reducing the demand of knowledge experts. A difference between our prototype system and the systems presented in the literature is, the possibility to increase accuracy of the generated ontology by selecting the more relevant terms and associations from the unstructured text and match them against the ontology design patterns. Also, a matching score is introduced to define the level of similarity between the terms extracted and the ontology design patterns. By setting a threshold value on the matching score, the relevant ontology design patterns are selected and used for the ontology construction process. The ontology construction framework used by the prototype system has been developed by the research group in Information Engineering of the School of Engineering, Jönköping University.

Les données disponibles sur le Web sont généralement de deux natures : (1) des données non structurées ou semi-structurées difficilement exploitables de manière automatique ou (2) des données structurées destinées à une utilisation particulière, difficilement réutilisables par d’autres applications. Le Web de données est une application du Web sémantique facilitant l’accès, le partage et l’alignement des données. Il existe actuellement de très nombreuses données disponibles sur le Web, mais qui ne sont pas publiées en suivant les principes du Web de données liées. Elles nécessiteraient d’être transformées en bases de connaissances. Nous proposons une méthodologie innovante qui permet de transformer plusieurs sources simultanément et non séquentiellement. Cette méthodologie permet la fusion de plusieurs sources de données orientée par des patrons de conception du domaine. Notre méthodologie spécifie la modélisation attendue du domaine en définissant la partie haute d’un module ontologique. Une chaîne de processus enrichit ce module par des éléments issus des sources : transformation syntaxique des sources, alignement, identification des éléments équivalents pour construire des candidats, calcul de score de confiance des candidats, filtrage des candidats. Notre travail part de l’hypothèse suivante : si un élément apparaît dans plusieurs sources, alors la possibilité qu’il appartienne au domaine d’étude est accrue. Nous avons défini différentes fonctions de calcul de la confiance consensuelle d’un candidat en mettant en évidence plusieurs caractéristiques comme le consensus entre sources ou la connectivité entre éléments d’un même candidat. Nous posons une deuxième hypothèse : un élément ne doit apparaître que dans un seul candidat pour obtenir une modélisation correcte. Cette hypothèse nous amène à définir la notion d’incompatibilité entre candidats. Nous pouvons considérer alors l’extraction des candidats qui ne partagent pas d’éléments, ce qui permet de faciliter le travail de validation. Pour évaluer nos propositions, nous avons mené trois expérimentations. La première a porté sur le domaine de la classification taxonomique des blés. Cette expérimentation nous a permis d’analyser la qualité des candidats générés avec l’aide de trois experts du domaine. La deuxième expérimentation a porté sur le même domaine et nous a permis de valider le temps gagné par un expert lors de la validation des candidats en considérant les incompatibilités. Pour la dernière expérimentation nous avons utilisé les données d’une campagne d’évaluation de systèmes d’alignements. Nous avons adaptés ces données pour évaluer la génération de candidats et la définition du score de confiance sur un grand jeu de données. Nous proposons une implémentation de cette proposition dans un outil réutilisable et paramétrable : Muskca. Celui-ci permet la fusion multi-sources pour la génération d’une base de connaissances consensuelle. L’application de nos travaux dans le domaine de l’agriculture nous a permis de constituer une base de connaissances sur la taxonomie des plantes. Cette base de connaissances permettra la représentation d’observations des attaques des agresseurs sur les cultures, ainsi que les techniques de traitement des agresseurs. Cette base de connaissances permettra de publier les données disponibles mais aussi d’annoter les nombreux documents mobilisables pour faire évoluer les pratiques agricoles. / The data available on the Web are generally of two kinds: (1) non structured data or semi structured data, which are difficult to exploit automatically; or (2) structured data, dedicated to a specific usage, which are difficult to reuse for a different application. The Linked Open Data is a Semantic Web application facilitating access, share ability and alignment of data. There are many data available on the Web, but these are not always published using the Linked Open Data theory and thus need to be transformed into knowledge bases. An innovative methodology is proposed in this work: one that transforms several sources simultaneously, not sequentially. This methodology merges several data sources oriented by domain design patterns and defines the expected domain representation using the upper part of an ontological module. A process chain enriches this module with elements from the sources: syntactic transformation of the sources, alignment, identification of equivalent elements for the construction of candidates, computation of the candidates’ trust scores and candidate filtering. This work is based on the following hypothesis: if an element appears in several sources then the possibility that it belongs to the studied domain is increased. Several functions were defined in order to compute the consensual trust score of a specific candidate by bringing out such characteristics as the consensus between the sources or the connectivity between the elements within a given candidate. A second hypothesis is put forward: to obtain a valid design, an element must be part of one candidate only. This hypothesis resulted in the definition of the notion of incompatibility between the candidates. The extraction of the candidates that do not share elements can then be considered, which made the experts’ validation task easier. To evaluate the proposals, three experiments were conducted. The first one dealt with the taxonomic classification of wheat. With the assistance of three experts, this experiment made for the analysis of the validation of the generated candidates. The second experiment, still in the same domain, lead to the evaluation of the time an expert saved using the notion of incompatibility during the validation of the candidates. As for the last experiment, the data from an evaluation campaign of alignment systems were used. These data had to be adapted to evaluate the generation of the candidates and the definition of the consensual trust score on a large data set. These three proposals were implemented in a new reusable and configurable tool: Muskca. This tool allows a multi-source fusion for the generation of a consensual knowledge base. This methodology was applied to agriculture, which allowed the creation of a knowledge base on plant taxonomy. The knowledge base will be used to represent the observations of pest attacks on crops along with pest treatment techniques. Not only will this knowledge base help the publication of the available data but it will also allow the annotation of the various documents that will be used, so as to improve agricultural practices.