Research on natural language place name query method
To address the issues of inaccurate understanding of the query intent,incorrect and missed search of place names during the query process,a natural language based place name query algorithm was proposed based on knowledge graph in this paper.Based on place name data for statistical classification,a corresponding map of place name knowledge was established,providing data support for queries;On the basis of analyzing the characteristics of common spatial query languages,an effective parsing scheme for natural language based spatial queries was proposed,which improved the intelligence of queries;A composite place name semantic similarity algorithm was proposed for place name abbreviations,aliases,and non-entered words,which solved some of the problems of errors and omissions in place name queries.Experimental results showed that the proposed query method could understand the query statements proposed by users in most cases and return more accurate results to users.
knowledge graphnatural language processingsemantic similarityplace name disam-biguation
NETL
NSTL
万方数据
