Customs declaration commodity intelligent classification method based on historical data mining

Abstract

The invention provides a customs declaration commodity intelligent classification method based on historical data mining. The method comprises the steps: analyzing the relation between a customs declaration commodity category code and the name, specification and model of a commodity, and describing a customs declaration commodity classification problem; preprocessing the historical information ofcustoms declaration commodities to remove useless parts of speech; designing an inverted index and a search algorithm for judging the first four codes of the commodity; performing feature selection onthe preprocessed commodities based on word frequency, and constructing a feature matrix based on a one-hot method; and based on the characteristics of different types of commodities, constructing a commodity classification model by adopting a decision tree algorithm; utilizing the classification model to classify customs declaration commodities, and obtaining a classification result, namely commodity codes. According to the invention, the category of the commodity can be well judged. The commodity can be effectively classified, and the commodity code based on the HS classification directory can be obtained. The method has high generalization performance. The customs clearance efficiency of enterprises can be improved. Trade risks caused by wrong classification of commodities are reduced.

Description

Technical field: [0001] The invention relates to the field of classification of customs declaration commodities, in particular to an intelligent classification method of customs declaration commodities based on historical data mining. Background technique: [0002] The HS ("Harmonized System") classification catalog adopted by my country's customs divides commodities into 21 categories and 97 chapters, and the chapters are further divided into items and subheadings. Based on the "Harmonized System", the total number of tax items in my country's "Tariff" has reached 8,547 (the 2017 edition of the "Tariff"). The traditional process of commodity classification is completed by professional classifiers. When faced with a large number of commodities that need to be classified, there are The shortcomings of long time-consuming and low efficiency, so according to the characteristics of the goods, automatically and intelligently classify the customs declaration goods and give the code...

AI Technical Summary

Problems solved by technology

Although the decision tree classification algorithm has been applied to some extent in other fields, there is still a lack of a complete and effective intelligent classification method based on decision trees in the field of customs declaration commodity classification.

Images