Method and system for obtaining national economy manufacturing industry corresponding to commodities based on machine learning

Abstract

The invention relates to a method and system for obtaining the national economy manufacturing industry corresponding to commodities based on machine learning. The method comprises the following steps:normalizing thousands of specific commodities into the classification of the bottommost layer in a commodity knowledge graph database established by a manufacturing industry system in a national standard commodity system through a machine learning algorithm; obtaining a mapping relationship between the bottommost commodity classification and the bottom classification of the national economy manufacturing industry by using a vector comparison algorithm; iteratively training a commodity classification prediction model capable of reflecting the mapping relationship from the commodity to the national economy manufacturing industry by utilizing the commodity classification and the mapping relationship at the bottommost layer. Therefore, the obtained commodity classification prediction model iswide in commodity coverage, completely covers thousands of minimum types designed by national standards, utilizes the most advanced deep learning model, and is excellent in prediction effect. According to the commodity classification prediction model, the commodity and the national economy manufacturing industry are successfully associated, and a good foundation is laid for other related applications.

Description

technical field [0001] The invention belongs to the field of Internet e-commerce, and in particular relates to a method and system for obtaining commodities corresponding to national economic manufacturing industries based on machine learning. Background technique [0002] With the rapid development of the Internet, more and more commodities are sold through e-commerce. It is more important to quickly and accurately classify the commodities sold. Commodity classification is to facilitate the needs of consumers to purchase, select appropriate classification signs for commodities, and systematically classify them step by step. However, most of the existing classification methods are based on statistics or simple machine learning techniques, so the classification effect obtained is not good, or the number of classifications that can be predicted is limited, the classification effect is not good, and the application after classification is single. Contents of the invention ...

AI Technical Summary

Problems solved by technology

[0003] Most of the current existing classification methods for the above existing technologies are based on statistics or simple machine learning techniques, so the classification effect obtained is not good, or the number of classifications that can be predicted is limited, the classification effect is not good, and after classification Applying a single technical problem, the present invention provides a method based on machine learning to obtain commodities corresponding to the national economic manufacturing industry. The method uses machine learning algorithms to associate commodities with national economic manufacturing industries. The commodity coverage is wide, and the deep learning model is used to , the prediction effect is good, laying a good foundation for related applications

Images