A user appeal analysis method and device

Abstract

The present application provides a user appeal analysis method and device, the method includes: using a preset word segmentation part-of-speech tagging model to perform word segmentation and part-of-speech tagging processing on user comment information, using a dependency relationship prediction model to perform dependency analysis processing, and using a preset The proposed classification model predicts the corresponding probability of candidate core words, and finally selects the candidate core words corresponding to the highest probability as the core words of user comment information. Since each model is trained according to product business keywords, it can be well adapted to the Analysis of user demands of product business. Because, when predicting the probability of candidate core words, the actual relationship between the words contained in the user comment information and the product business is fully considered, so the user's actual demands for the product business can be fully mined. It can be seen that the technical solution provided by this application can automatically mine user appeals, and its mining efficiency and accuracy are much higher than the manual reading methods in the prior art.

Description

technical field [0001] The present application relates to the technical field of Internet data processing, and in particular to a user appeal analysis method and device. Background technique [0002] With the popularization of the Internet and the rapid development of e-commerce, a large amount of consumer review information on products is stored on the Internet, which includes consumers' positive or derogatory comments on product performance or functions. Merchants can obtain feedback from consumers in a timely manner by tracking these evaluation information, thereby improving products. [0003] Therefore, the research on opinion mining of user comment information emerges. How to quickly and effectively mine user appeals from comment information has become a research hotspot in comment information processing. However, at present, user appeals are often mined by manually reading user comments. However, in the face of a large amount of unstructured comment information on the...

AI Technical Summary

Problems solved by technology

However, at present, user appeals are often mined by manually reading user comments. However, in the face of a large amount of unstructured comment information on the Internet, mining user appeals by manual reading is a time-consuming and laborious process.

Images