A framework and implementation for qualitative geographic information retrieval

Qualitative geographic information retrieval (QGIR) is proposed in this paper. It is semantic matching oriented and based on the qualitative representation. Qualitative reasoning is used to get the relevance between documents and queries. Compared to the traditional geographic information retrieval (GIR), QGIR uses qualitative representation for both themes and place names. Semantic spatial relationships are taken into account. The qualitative representation is closer to human spatial recognition and can handle natural language better. A system is built to verify the qualitative theory and to compare the qualitative method with traditional GIR. The experiment in the mineral filed shows that QGIR can have better searching results and better fulfill people´s geographic information retrieval request.