Geographic Information Retrieval based on multiple formulations and search engines

This work is related to the area of Geographic Information Retrieval (GIR). In this work, a GIR system has been designed, developed and evaluated, applying different techniques of Natural Language Processing (NLP), several information retrieval engines and using external geographic knowledge resources. The developed system is modular, allowing the integration of other modules or knowledge resources. The main novelty of the system is to use an architecture based on creation of different subqueries from the original query, the use of three textual information retrieval tools and different fusion techniques of document lists, achieving improvements in variability and coverage of the documents retrieved.