A Method for Structural Extraction of Location Concept Based on Salient Structure of Location Description

Abstract

The invention discloses a position description based position concept structuralization extracting method for significance structures. Firstly, structure body modeling is performed based on position description, position concept significance structures are summarized, significant syntax languages are customized, and corresponding categories are read and generated during program initialization; then, position structure concept bodies are automatically mapped to generate rule classes, each rule class correspond to one position entity concept body, and finally rule sets are generated; then, the rule sets are classified and are re-divided into a plurality of rule groups to form rule retrieval dependence diagrams; finally, any natural language text position description statements and original language text position description statements serve as inputs, matching is performed by utilizing a rule matching algorithm based on position concept significance structures, and structuralization extraction of position information is completed through grading and sorting. The significance structures are used for supplementing and optimizing knowledge representations of semantic information of placename structures and are conductive to improvement of matching result accuracy.

Description

technical field [0001] The invention belongs to the technical field of natural language location extraction, in particular to a location concept structured extraction method based on location salience description. Background technique [0002] Multi-source heterogeneous place name data has increased in a large amount. In order to share and integrate location data from different sources and structures, and to accurately obtain the prepared query results based on location name retrieval, it is necessary to start from the perspective of people's cognitive habits and analyze a large number of standard and non-standard data. The place name description is analyzed, the structural characteristics and spatial relationship of the location name are summarized, the semantic information of the place name structure is conceptually and formally expressed, and the structural modeling of the location concept is realized. On this basis, in order to identify the location name and location rel...

AI Technical Summary

Problems solved by technology

[0003] Due to possible expression errors and incomplete knowledge bases in the actual location description, among the place name description objects obtained by the location concept matching method, many programs are intermediate results or are not significant enough to represent the location concept compared to other results.

General location concept matching can only match string information, it is difficult to flexibly integrate the semantic relationship between place name entities and relational words, and it also ignores people's cognition of place names and their description habits of place names

In addition, when the number of rules is large and matches fine-scale place names, the matching efficiency is very low, so it is necessary to define the saliency structure of place name descriptions to perform coarse filtering on the unsalient matching results generated by location concept matching

Images