Full text
| |
| Author(s): |
Caio Saraiva Coneglian
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | Marília. 2017-03-13. |
| Institution: | Universidade Estadual Paulista (Unesp). Faculdade de Filosofia e Ciências. Marília |
| Defense date: | 2017-02-17 |
| Advisor: | José Eduardo Santarem Segundo |
| Abstract | |
The evolution of Information and Communication Technologies has caused an exponential increase in the production and dissemination of data on the Internet. Among these data, we include the scientific production that is now in a moment of transition, in which the documents are no longer only published in printed books and magazines, and are now spread throughout the network. From this, a new technology has emerged called digital repositories, in which documents are stored in various formats. Along with the emergence of digital repositories, the challenge of recovering these documents has grown in an efficient way, i.e., how the machine can understand what the user is looking for, to provide the documents that this user needs. In this context, the Semantic Web came about in order to enable computers to understand the context in which the information created by the users meet, having in their tools the basis to make such a real purpose. In the context of Digital Repositories, this research aims to improve the Information Retrieval process in these informational environments through the use of the concept of semantic representations in the use of domain ontologies, which allows greater adherence at the intersection between bibliographic items and Information needs of users. In order to achieve these objectives, a quantitative methodology was used, in which a digital repositories interoperability model was created, using Semantic Web concepts and technologies to contextualize the search domain performed by the user. As results, it was verified that the relationships of the ontologies can be extracted efficiently by means of a SPARQL query engine, that is able to locate a term in the ontology, as well as to extract the relations of this term. In addition, we have outlined the actions that OWL properties must possess, in the context of Information Retrieval, so that the relationships that a search term has against an ontology can be identified more precisely, allowing the generation of a new Search expression, containing a greater number of arguments. Another result obtained concerns interoperability in digital repositories, which made it possible to identify the integration of the retrieval of the metadata of the digital repositories documents and the tool dealing with semantic issues. The work proposed the interactivity in the choice of informational sources, in which the user could choose the repositories in which the search would be carried out, as well as register a repository, if it had not previously been used. It is concluded that the insertion of semantics in Information Retrieval processes can occur through the proposed model, which is based essentially on the technologies and concepts of the Semantic Web, especially the ontologies, as an artifact capable of explaining the context in which the terms occur. (AU) | |
| FAPESP's process: | 15/01517-2 - Definition of information retrieval model for digital repositories based on semantic web technologies |
| Grantee: | Caio Saraiva Coneglian |
| Support Opportunities: | Scholarships in Brazil - Master |