Geographic element querying and extending method based on semantic similarity

Abstract

The invention relates to a geographic element querying and extending method based on semantic similarity, wherein the concept similarity is calculated through the shortest distance between two concept nodes of a geographic ontology while considering node depth, node density impact factor, common attribute proportion and public ancestors node density and other impact factors. In addition, the characteristics for differing geographic ontology from common domain ontology, namely the spatial relationships among the geographical concepts, are considered fully, for example, position relation, topological relation and metric relation and etc.; the characteristics, as important impact factors, are provided with different weights according to the different spatial relationships among the concepts, so that the impacts of spatial relationships to the semantic similarity are quantized. The geographic element querying and extending method based on semantic similarity provided by the invention has the advantages that the geographic element querying and extending purpose based on the semantic similarity of geographic ontology is achieved while fully considering and selecting the various of impact factors, the information searched by the users can be quickly and accurately obtained, the querying method based on simple morphology matching is changed, and the efficiency of querying is improved.

Description

technical field [0001] The invention belongs to the field of geospatial information systems, and in particular relates to a method for extending geographic element query based on semantic similarity of geographic ontology. Background technique [0002] "Human beings are being overwhelmed by data, but human beings are poor in knowledge." With the rapid development of computer and Internet technology, the amount of information on the network has increased dramatically. Facing the huge information resources, people feel at a loss and get lost in the huge Internet. In China's information network, "information overload" and "information trek" have become one of the most talked about topics. Facing the vast amount of information, a hot issue in the field of information technology is how to help users retrieve and obtain really useful information with high quality. However, the current practical information retrieval system is mainly based on manual classification and keyword matc...

AI Technical Summary

Problems solved by technology

[0006] (2) The fuzziness of geospatial information description

One aspect is the use of names that cannot be pinpointed when describing geospatial information

On the other hand, there is ambiguity in the description of geospatial information

For example, some place names in different regions have the same name, or the place name and the location name are the same, these names are very easy to cause confusion in the information retrieval of the traditional word form matching method, and cause obstacles to the retrieval of geospatial information

[0007] (3) Traditional information retrieval is mainly realized by keyword morphological matching, and the existence of heterogeneous information such as polysemous words and synonyms makes the defects of this morphological matching method increasingly obvious

However, this method is more dependent on the corpus used for training, the amount of calculation is large, and the calculation method is complicated. In addition, it is greatly disturbed by data sparseness and data noise, and sometimes obvious errors may occur.

The method based on the dictionary is simple and effective, and it is more intuitive and easy to understand. However, the results obtained by this method are greatly affected by people's subjective consciousness, and sometimes cannot accurately reflect objective facts.

In addition, this method more accurately reflects the similarities and differences in semantics between words, and takes less consideration of the syntactic and pragmatic features between words

Although the third type of method takes into account factors such as node depth, node density, and attribute information that affect the semantic similarity of concepts, it only considers the basic attribute information and ignores the influencing factor of the spatial relationship between geographic concepts. is the most influential factor on semantic similarity in geographic ontology, so it is difficult to apply it in calculating the similarity of geographic features

Images