Document retrieval using fuzzy related geographic ontologies

Many documents stored in digital libraries and document database include geographic references within their texts typically by means of place names. These geographic references can be associated to knowledge to help improve related retrieved documents. For example, in a meteorological document collection the climate knowledge can be associated to the geographic references allowing the retrieval of relevant climate or geographic related documents to an initial user query. In order to explore these issues this work describes a framework to encode a geographic knowledge base composed of multiple related ontologies whose relationships are expressed as fuzzy relations. Each ontology represents a distinct area of the knowledge domain related to geographic references. This knowledge organization is used in a fuzzy method to expand the user initial query. Each ontology can be represented independently as well as their relationships. The fuzzy query expansion method is tested with the Apache Lucene search engine improving the precision measure.