Fusion method based on diversified geographic information points

Abstract

The invention relates to a fusion method based on diversified geographic information points, and belongs to the technical field of geographic information processing. According to the method, the geographic information data matching is performed through finding a certain identical region in a map with different geographic reference coordinate systems; whether geographic information of different maps has the identical information points or not is found; and the method relates to the field of geography and information processing. The method comprises the following steps of defining the similarity degrees of geographic names in a plurality of geographic information points; obtaining a plurality of similarity degree values through calculation according to the result after the text segmentation and the text name entity recognition; then, whether the geographic names of a plurality of geographic information points are similar or not is judged according to the final result condition of permutation and combination and additive operation; if the geographic names are not similar, the matching is completed, and the geographic names of the plurality of geographic information points are different; and if the geographic names are similar or the condition is uncertain, the matching between latitude and longitude distances of the plurality of geographic information points is performed.

Description

technical field [0001] The invention relates to a fusion method based on multiple geographic information points, belonging to the technical field of geographic information processing. Background technique [0002] In recent years, with the large-scale development of the network, the number of various information resources has increased sharply, and the popularity of computers for various natural language processing applications such as geographical names and coordinate information. People need a fast and accurate method to calculate multiple geographic information. Information similarity between information points. In particular, the calculation of short text similarity plays a very important role, and its application can greatly improve the accuracy of identifying multiple geographic information points. In addition, the latitude and longitude information in the map is very convenient for the query of adjacent areas, and the fusion method of establishing geographic informat...

AI Technical Summary

Problems solved by technology

But for short texts, because similar short texts do not necessarily have the same words, the flexibility of natural language allows people to express the same meaning through different wordings, so the existing similarity calculation methods are not effective. Good, e.g. poor performance in data handling of geolocation names

In addition, in the traditional text similarity calculation process, stop words are usually removed from the expression of the text, such as "的", "是", "中", etc., but in the process of processing short texts, because they contain some sentence structure information, so some stop words cannot be ignored, which makes the traditional text similarity calculation method can not get good results in the process of text similarity calculation for short text

[0005] For the calculation of text similarity, the commonly used method for calculating the similarity between texts is the cosine similarity method, which regards the text as a point in space and expresses it as a vector form, and uses the angle between the vectors to Quantitatively calculate the similarity between texts, this method does not consider the feature words with the same semantics between texts, and cannot fully reflect the semantic similarity between texts

Therefore, there are a lot of shortcomings and omissions in the existing similarity calculation methods

Images