Character approximation degree retrieval analysis method based on big data feature recognition

Abstract

The invention discloses a character approximation degree retrieval analysis method based on big data feature recognition, which comprises the following steps of: constructing a character approximation retrieval platform, carrying out preliminary approximation matching on characters to be subjected to approximation retrieval to obtain successfully matched candidate characters, and screening the candidate characters without taking radicals of target characters as a retrieval basis. Therefore, the range of candidate characters of approximate retrieval corresponding to the target characters is effectively expanded, the characters to be approximately retrieved are subjected to overall structure, character coordinate and composition stroke feature recognition, and then the features serve as the similarity analysis basis for similarity analysis of the candidate characters and the characters to be approximately retrieved. Therefore, the approximation degree analysis results are synthesized to obtain the comprehensive approximation retrieval result corresponding to the character to be subjected to approximation retrieval, the approximation degree analysis basis is diversified, and the defect that in an existing character approximation degree retrieval analysis method, due to the fact that the approximation degree retrieval analysis basis is too single, the approximation degree analysis result is too one-sided is greatly overcome.

Description

technical field [0001] The invention belongs to the technical field of text proximity retrieval, and in particular relates to a text proximity retrieval and analysis method based on big data feature recognition. Background technique [0002] With my country's entry into WTO, enterprises are participating in increasingly fierce competition worldwide. As an intangible capital of enterprises, trademarks play an important role in the survival and development of enterprises. Under such circumstances, companies from all walks of life are scrambling to register their corporate trademarks first. For trademarks registered solely in words, due to the characteristics of the shape and structure of Chinese characters, there are many similar characters in Chinese characters, which leads to the frequent occurrence of similar trademarks in word-registered trademarks. The similarity examination of similar trademarks requires the use of text similarity retrieval and analysis methods in the p...

AI Technical Summary

Problems solved by technology

[0003] However, most of the current text similarity retrieval and analysis methods use the radical components of the text to be retrieved as the basis for similarity retrieval and analysis. This kind of similarity retrieval and analysis method has a single basis for retrieval and analysis. On the one hand, it only Texts with the same radical components as the text to be approximated can be retrieved. For some texts that have different radicals from the text to be approximated but have similar overall shapes, the retrieval is usually missed, resulting in limited retrieval scope; In the process of analyzing the similarity between the text and the text to be retrieved, the similarity analysis is performed only by the components of the text, resulting in relatively one-sided analysis results and low accuracy, making it difficult to achieve comprehensive and accurate text similarity retrieval analysis

Images