Bayesian model-based commodity code classification method and system

Abstract

The invention provides a Bayesian model-based commodity code classification method and system. Aiming at the problems of sparsity and context loss caused by short text features of commodity name information, the method and system adopt a synonym extension technology for word-segmented sparse words, and increase the word synonym degrees through an external search engine technology; and then, a Bayesian model is adopted to calculating condition probabilities between word sequences and code categories of commodity names so as to obtain a Bayesian code classification mode, and a trained model is used for predicting commodity code categories of input commodity names.

Description

technical field [0001] The invention belongs to the field of text information analysis and processing, and more specifically relates to a Bayesian model-based commodity code classification method and system. Background technique [0002] In tax control machines, computer software with the function of issuing invoices and other equipment and applications, it is generally necessary to retrieve and assign a tax commodity code corresponding to the commodity name based on the commodity name. The general tax commodity coding structure is composed of major categories, subcategories, etc. There are more than 4,300 categories at the most detailed level, and the commodity coding classification rules mainly stipulate the types of commodities, while the brand or other categories of commodities Properties are not specified. In the process of invoicing, if you rely on manual commodity classification, you need to judge which category belongs to the general tax commodity code according to ...

AI Technical Summary

Problems solved by technology

[0004] Relying on the fuzzy query and keyword query of the database can solve the problem of assigning codes to some product names, but because the product names vary widely, and due to the heterogeneity of Chinese characters, the accuracy of this method is not high

In addition, the product name presents short text characteristics, and the lack of contextual semantics leads to its semantic sparsity and semantic sensitivity, which makes it impossible to apply traditional statistical models

Images