Ontology integration is a process of matching and merging two ontologies for reasons such as for generating a new ontology, thus creating digital services and products. Current techniques for ontology integration, used for information and knowledge integration, are not powerful enough to handle the semantic and pragmatic heterogeneities. Because of the heterogeneities, the ontology matching and integration have shown to be a complex problem, especially when the intention is to make the process automatic. This thesis addresses the problem of integrating heterogeneous ontologies, first, by exploring the context of ontology integration, secondly, by building a context knowledge base, and thirdly, by applying the context knowledge base. More specifically, the thesis contributes a context knowledge base method for ontology integration, CKB-OI method, which contains: 1) A method of building a context knowledge base by extracting context and contextual information from ontologies in an ontology repository to improve ontology integration. 2) A method of refining the result of ontology integration with the help of the context knowledge base and expanding the context rules in the context knowledge base. In the first method, the context of the ontology integration is identified by examining the content and metadata of the integrated ontologies. The context of an ontology integration contains the information describing the integration, such as the domain of ontology, the purpose of ontology, and the ontology elements involved. Context criteria, such as the metadata of ontologies and the element of ontologies in the repository, are used to model the context. The contextual information is extracted and integrated from ontologies in an ontology repository, using an ontology integration process with non-violation check. With the context and the contextual information, a context knowledge base is built. Since this is built by reusing ontologies to provide extra information for new ontology integration in the same context, it is quite possible that the context knowledge base will improve the earlier ontology integration result. A method for identifying the domain of an ontology is also proposed to help in building and using the context knowledge base. Since the method considers the semantic and pragmatic heterogeneities of ontologies, and uses a light-weight ontology representing a domain, this work increases the semantic value of the context knowledge base. In the second method, the context knowledge base is applied to the result of an ontology integration process with a non-violation check, which in turn results in an ontology intersection. The contextual information is searched for and extracted from the context knowledge base and then applied on the ontology intersection to improve the integration result. The ontology non-violation check integration process is adjusted and adopted in the method. Moreover, the context knowledge base is expanded with perspective rules, with which the different views of ontologies in a context are preserved, and reused in future ontology integration. The results of the CKB-OI methods are: 1) a context knowledge base with rules that consider semantic and pragmatic knowledge for ontology integration; 2) contextual ontology intersection (COI) with the refining result compared to the ontology intersection (OI), and 3) an extended context knowledge base with the different views of both ontologies. For evaluation, ontologies from the Ontology Alignment Evaluation Initiative (OAEI) and from ontology search engines Swoogle and Watson have been used for testing the proposed methods. The results show that the context knowledge base can be used for improving heterogeneous ontologies integration, hence, the context knowledge base provides semantic and pragmatic knowledge to integrate ontologies. Also, the results demonstrate that ontology integration, refined with the context knowledge base, contains more knowledge without contradicting the ontologies involved in our examples. / Ontologi-integration är en process för att matcha och sammanfoga två ontologier för att t.ex. generera en ny ontologi, och därmed skapa digitala tjänster och produkter. Aktuella tekniker för ontologi- integration, som används för information och kunskapsintegration, är inte tillräckligt kraftfulla för att hantera semantiska och pragmatiska heterogeniteter. På grund av heterogeniteter, har ontologi- matchning och -integration visat sig utgöra ett komplext problem, särskilt när avsikten är att göra processen automatisk. Denna avhandling behandlar problemet med att integrera heterogena ontologier; för det första genom att undersöka kontexten för ontologi-integrationen, för det andra genom att bygga en kunskapsbas för kontexten, och för det tredje genom att tillämpa denna kunskapsbas. Mer specifikt bidrar avhandlingen med CKB-OI-metoden för ontologi-integration, vilken innehåller: 1)      En metod för att bygga en kontextkunskapsbas, genom att extrahera sammanhang och kontextuell information från ontologier i ett ontologi-förvar för att förbättra ontologi-integrationen. 2)      En metod för att förfina resultatet av ontologi-integration med hjälp av kontextkunskapsbasen och för att utöka kontextreglerna i kunskapsbasen. I metod nr. 1 identifieras kontexten genom att undersöka innehållet och metadata för de ontologier, som ska integrereras. Kontexten innehåller information som beskriver integrationen, till exempel domän och syfte för varje ontologi, samt element som ingår i respektive ontologi. Kontexten  modelleras med kriterier, såsom metadata och element för ontologierna i förvaret. Den kontextuella informationen extraheras och integreras med användning av en integrationsprocess med icke-överträdelsekontroll. Kontextkunskapsbasen byggs utav kontext samt kontextuell information. Eftersom kunskapsbasen är byggd av återanvända ontologier för att ge ytterligare information till ontologi-integrationen inom samma kontext, så är det mycket möjligt att kontextkunskapsbasen kommer att förbättra det tidigare integrationsresultatet. En metod för att identifiera domänen för en ontologi föreslås också, för att hjälpa till att bygga och använda kontextkunskapsbasen. Eftersom metoden tar hänsyn till de semantiska och pragmatiska heterogeniteterna hos ontologier, och använder en enkel ontologi för att representera en domän, så ökar detta arbete det semantiska värdet av kontextkunskapsbasen. I metod nr. 2 tillämpas kontextkunskapsbasen på resultatet av en ontologi-integrationsprocess med icke-överträdelsekontroll, vilket i sin tur resulterar i ett ontologisnitt. Den kontextuella informationen extraheras från kontextkunskapsbasen och appliceras sedan på ontologisnittet för att förbättra integrationsresultatet. Icke-överträdelsekontrollen i integrationsprocessen justeras och används på nytt. Dessutom utökas kontextkunskapsbasen med perspektivregler, med vilka de olika vyerna av ontologier i en gemensam kontext bevaras och återanvänds i framtida ontologi-integrationer. Resultaten av CKB-OI metoden är: 1) en kontextkunskapsbas med regler som avser semantiska och pragmatiska kunskaper om en ontologi-integration; 2) ett kontextuellt ontologisnitt (COI) med ett förfinat resultat jämfört med ontologisnittet (OI) och 3) en utökad kontextkunskapsbas med olika vyer av båda ontologier. För utvärderingen har ontologier från Ontology Alignment Evaluation Initiative (OAEI) samt ontologisökmotorerna Swoogle och Watson använts för att testa de föreslagna metoderna. Resultaten visar att kontextkunskapsbasen kan användas för förbättring av heterogena ontologi-integrationer. Följaktligen tillhandahåller kontextkunskapsbasen semantiska och pragmatiska kunskaper för att integrera ontologier. Dessutom visar resultaten att ontologi-integrationer, utökade med kontextkunskapsbaser, innehåller mer kunskap, utan att motsäga de ontologier som ingår i våra exempel. / <p>QC 20141017</p>

Ontologies are widely used in various domains, including biomedical research, to structure information, represent knowledge, and analyze data. The combination of ontologies from different domains is crucial for systematic data analysis and comparison of similar domains. This process requires ontology composition, integration, and alignment, which involve creating new classes by reusing classes from different domains, aggregating types of ontologies within the same domain, and finding correspondences between ontologies within the same or similar domain. This thesis presents use cases where we applied ontology composition, integration, and alignment of phenotype ontologies, and evaluated the resulting ontologies and alignment. First, we analyzed a large aging dataset of inbred laboratory mice, using Mouse Anatomy and Mouse Pathology ontologies. Second, we integrated phenotype ontologies for human and model organism phenotypes to enable comparisons of phenotypes between and within individual species. We developed Pheno-e, an extension of PhenomeNet. We identified novel abnormal anatomical classes for fly phenotypes, allowing the annotation of fly genes that were not annotated before. We demonstrate the distinct contributions of each species' phenotypic data to detecting human diseases using Pheno-e, and show that mouse phenotypic data contributes the most to the discovery of gene--disease associations. This work could guide the selection of model organisms when building methods to find gene-disease associations. Additionally, we refined class definitions in phenotypic ontologies, specifically targeting cell cardinality phenotypes. This representation resolved incorrect inferences in the utilized ontologies, enabling accurate interpretation of phenotypic descriptions. Our findings reveal that this correction enhances gene-disease prediction for diseases associated with cardinality phenotypes. Third, we introduce a novel neural-symbolic method that combines logic fundamentals with machine learning for ontology alignment. This method begins with symbolic representation, followed by iterative neural learning for alignment and symbolic representation consistency checking and reasoning, and back to neural learning. We demonstrate that our system generates noncontroversial alignments first and these alignments are coherent with respect to OWL EL. This novel method can pave the way for more accurate and efficient ontology-based methods, which can have significant implications for various semantic web applications.

Made available in DSpace on 2016-08-17T14:53:28Z (GMT). No. of bitstreams: 1 Suzane Carvalho dos Santos.pdf: 4549168 bytes, checksum: 85d08a343bc93d5bf241da9f6f02f5b4 (MD5) Previous issue date: 2014-08-18 / Ontologies are knowledge representation structures capable of expressing a set of entities of a domain, their relationships and axioms that are being used by modern knowledge based systems (KBS) in the decision making process. However, manual construction of ontology is expensive and subject to errors, thus a viable alternative is the automation of this process. Several techniques and tools have been developed to learn the different components of an ontology from textual sources, named concepts, hierarchies, instances, relationships, properties and axioms. However, these elements are generally acquired in a isolated manner. Due to the lack of approaches to acquire all the elements of an ontology jointly, there is a need to develop a process to make the reuse and the learning of each of the elements of an ontology in a synergistic manner. To attend this need, this work presents Apponto-Pro, an incremental learning process for populating application ontologies from textual information sources that is capable of generating a complete ontology through the integration of different techniques to generate isolated elements of an ontology. The process was evaluated through a case study that consisted in the automatic construction of Family_Law, an application ontology in the field of family law developed with Apponto-ProTool, a software tool to support Apponto-Pro that integrates the approaches that compound the whole process. This evaluation aimed to determine the effectiveness of the ontology constructed with Apponto-ProTool against an ontology manually built by a domain specialist and used as reference ontology. For this reason, the "precision"was calculated for the elements of the ontology automatically generated using the reference ontology. As a result it was found that in some cases the ontology developed with Apponto-ProTool tends to present more suitable results. / As ontologias são estruturas de representação de conhecimento capazes de expressar um conjunto de entidades de um dado domínio, seus relacionamentos e axiomas, sendo utilizadas pelos modernos Sistemas Baseados em Conhecimento (SBC) no processo de tomada de decisões. No entanto, a construção manual de ontologias é cara e sujeita a erros, sendo uma alternativa viável a sua construção de forma automática. Diversas técnicas e ferramentas têm sido desenvolvidas para aprender os diferentes componentes de uma ontologia a partir de fontes textuais, quais sejam conceitos, hierarquias, instâncias, relacionamentos, propriedades e axiomas. Entretanto estes elementos são, em regra, adquidiros de forma isolada. Devido à carência de abordagens que adquirem todos os elementos de uma ontologia de forma conjunta, surgiu a necessidade de desenvolver um processo que faça o reúso e a aprendizagem de cada um dos elementos de uma ontologia de forma completa. Atendendo a esta necessidade, este trabalho apresenta o Apponto-Pro, um processo incremental para o aprendizado e povoamento de ontologias de aplicação a partir de fontes de informação textuais capaz de gerar uma ontologia completa através da integração de diferentes técnicas que geram elementos da ontologia de forma isolada. O processo foi avalizado através de um estudo de caso que consistiu na construção automática da Family_Law, uma ontologia de aplicação no domínio do Direito da Família construída através da aplicação da ferramenta de software Apponto-ProTool, desenvolvida para dar suporte ao processo Apponto-Pro que integrou as ferramentas correspondentes as abordagens contidas no processo. Esta avaliação teve como objetivo verificar a efetividade da ontologia construída pela Apponto-ProTool em relação a uma ontologia construída manualmente por um especialista do domínio e utilizada como ontologia de referência. Para isso foi calculado o valor da medida "precision" para os elementos da ontologia construída utilizando a ontologia de referência. Como resultado verificou-se formalmente que em alguns casos a ontologia desenvolvida pela Apponto-ProTool tende a apresentar resultados mais adequados.

Cette thèse se situe dans le champ de la gestion des connaissances à l’aide de modèles ontologiques. Pour représenter les connaissances de domaine, nous avons conçu une ontologie hybride à deux niveaux : au niveau local, chaque groupe d’experts (du même métier) a construit sa propre ontologie, au niveau global une ontologie consensuelle regroupant les connaissances partagées est créée de façon automatique. De plus, des liens sémantiques entre les éléments de différentes ontologies locales peuvent être ajoutés.Nous avons construit un système d’aide pour guider les experts dans le processus de création de liens sémantiques ou mises en correspondance. Ses particularités sont de proposer des mesures de similarité en fonction des caractéristiques des ontologies à aligner, de réutiliser des résultats déjà calculés et de vérifier la cohérence des mises en correspondances créées.Par ailleurs, les ontologies locales peuvent être mises à jour. Cela implique des changements au niveau de l’ontologie globale ainsi que des mises en correspondances créées. De ce fait, nous avons développé une approche, adaptée à notre domaine pour gérer l’évolution de l’ontologie hybride. En particulier, nous avons utilisé la notion de versions d’ontologies afin de garder trace de toutes les modifications apportées au niveau des ontologies et de pouvoir revenir à tout moment à une version précédente.Nous avons appliqué notre travail de recherche à la géotechnique qui est un domaine complexe impliquant des experts de différents métiers. Une plateforme logicielle est en cours de réalisation et permettra de tester la faisabilité de nos travaux. / This thesis concerns the scope of knowledge management using ontological models.To represent domain knowledge, we design a hybrid ontology on two levels: In a local level, each experts’ group has designed its own ontology. In a global level, a consensual ontology containing all the shared knowledge is automatically created.We design a computer-aided system to help experts in the process of mapping creation. It allows experts to choice similarity measures relatively to the ontology characteristics, to reuse the calculated similarities and to verify the consistency of the created mappings.In addition, local ontologies can be updated. This involves modifications in the global ontology and on the created mappings. A relevant approach of our domain was developed.In particular, ontology versioning is used in order to keep a record of all the occurred modifications in the ontologies; it allows to return at any time a previous version of the hybrid ontology.The exploited domain is geotechnics which gathers various business experts. A prototype is in progress and currently does not still captures ontology evolution.

Le but essentiel de la thèse est de développer une ontologie juridique bien fondée pour l'utiliser dans le raisonnement à base des règles. Pour cela, une approche middle-out, collaborative et modulaire est proposée ou des ontologies fondationnelles et core ont été réutilisées pour simplifier le développement de l'ontologie. L’ontologie résultante est adoptée dans une approche homogène a base des ontologies pour formaliser la liste des règles juridiques du code pénal en utilisant le langage logique SWRL. / This thesis analyses the problem of building well-founded domain ontologies for reasoning and decision support purposes. Specifically, it discusses the building of legal ontologies for rule-based reasoning. In fact, building well-founded legal domain ontologies is considered as a difficult and complex process due to the complexity of the legal domain and the lack of methodologies. For this purpose, a novel middle-out approach called MIROCL is proposed. MIROCL tends to enhance the building process of well-founded domain ontologies by incorporating several support processes such as reuse, modularization, integration and learning. MIROCL is a novel modular middle-out approach for building well-founded domain ontologies. By applying the modularization process, a multi-layered modular architecture of the ontology is outlined. Thus, the intended ontology will be composed of four modules located at different abstraction levels. These modules are, from the most abstract to the most specific, UOM(Upper Ontology Module), COM(Core Ontology Module), DOM(Domain Ontology Module) and DSOM(Domain-Specific Ontology Module). The middle-out strategy is composed of two complementary strategies: top-down and bottom-up. The top-down tends to apply ODCM (Ontology-Driven Conceptual Modeling) and ontology reuse starting from the most abstract categories for building UOM and COM. Meanwhile, the bottom-up starts from textual resources, by applying ontology learning process, in order to extract the most specific categories for building DOM and DSOM. After building the different modules, an integration process is performed for composing the whole ontology. The MIROCL approach is applied in the criminal domain for modeling legal norms. A well-founded legal domain ontology called CriMOnto (Criminal Modular Ontology) is obtained. Therefore, CriMOnto has been used for modeling the procedural aspect of the legal norms by the integration with a logic rule language (SWRL). Finally, an hybrid approach is applied for building a rule-based system called CORBS. This system is grounded on CriMOnto and the set of formalized rules.