CAMPELO, CLAUDIO E. C.; http://lattes.cnpq.br/2042247762832979; CAMPELO, Cláudio Elízio Calazans.
Résumé:
Geographic Information Retrieval (GIR) is a recent research area which has become
very attractive. Regarding the fast growth of the amount of information available on the Web, the provision of simple and efficient retrieval tools to be used by different user needs is currently one of the most challenge in the field of Information Retrieval (IR). Geographic Web search engines are specializations of the standard Web search engines, which add to them the ability to identify the geographic context of Web resources (e.g., texts, images, movies) and to index them using the spatial features. Research on GIR may be categorized in a few subareas which have specific objectives: to develop methods to detect geographic references in Web resources; to elaborate representation models for these resources' geographic scope, that is, the set of places the documents are associated to; to build efficient data structures and algorithms for spatio-textual indexing; to conceive models for relevance ranking, which may combine both spatial and textual dimensions; and to design user interfaces and search mechanisms which allow to represent user queries by these dimensions and then to retrieve the most relevant documents. This work presents an elaborated model for a GIR system building, emphasize both geographic scope modeling and relevance ranking generation processes. In order to validate the proposed model, a geographic Web search engine prototype, called GeoSEn – GEOgraphic Search ENgine, was built. GeoSEn integrates the main functions contained in a geographic Web search engine.