An address similarity evaluation method

Abstract

The invention relates to an address similarity evaluation method, which belongs to the technical field of information processing. The present invention abandons the simple use of address text information or longitude and latitude information to determine the similarity of addresses, but innovatively divides address text information into six-segment word sets according to administrative regions, and assigns weights to word sets in a six-segment style, and then through The Jaro-Winler Distance algorithm calculates its similarity based on text information, calculates its similarity based on latitude and longitude through coordinate conversion and Pythagorean theorem, and calculates its similarity based on word sets through the cosine theorem. The three similarities are reflected from different aspects The degree of similarity of the addresses is calculated, and finally the three calculated similarities are fused to obtain the final similarity. Compared with the prior art, the present invention mainly solves the lack of accuracy caused by blind pursuit of efficiency in the prior art, and devotes itself to increasing the accuracy of address similarity evaluation currently relying on computers.

Description

technical field [0001] The invention relates to an address similarity evaluation method, which belongs to the technical field of information processing. Background technique [0002] Geographic information systems are increasingly used frequently in current business systems. For example, some group-buying apps will search for surrounding business information based on the user's location; or when developers make new maps, they need to refer to existing maps for comparison of specific addresses, etc. [0003] At present, the evaluation methods for address similarity are mainly divided into two categories: one is to use address text information as input, and use string similarity algorithms, such as Levenshteit Distance, to measure the similarity of address text; the other is to use longitude and latitude information comparison to obtain their similarity. However, these two types of methods have their own defects. If you directly use the string similarity algorithm to compare...

AI Technical Summary

Problems solved by technology

However, these two types of methods have their own defects. If you directly use the string similarity algorithm to compare the address text, once there is an abbreviated form in the address text, this method will have a large error; if you only rely on latitude and longitude information to determine, it is also very unreliable. method, once the specific addresses of multiple existing maps are compared, because each map has its own unique coordinate system, multiple conversions will make the accuracy inaccurate and lead to large errors in the evaluation results

Images