User abnormal comment detection method and system based on hierarchical multi-channel attention

Abstract

The invention relates to the field of natural language processing, in particular to a user abnormal comment detection method based on hierarchical multi-channel attention, which comprises the following steps: collecting user comment data in real time, inputting the collected user comment data into a trained user comment data detection model, and judging whether a comment is abnormal or not; according to the method, tedious feature engineering design in detecting the abnormal comments of the user based on a machine learning method is avoided, the features of the user comment text are analyzed,and the problem that global semantics cannot be captured or cannot be fully captured by a single deep neural network model and a single-layer attention mechanism is solved.

Description

technical field [0001] The invention relates to the field of natural language processing, in particular to a method and system for detecting user abnormal comments based on hierarchical multi-channel attention. Background technique [0002] With the rapid development of the Internet, the scale continues to expand, and the number and influence of online reviews are also increasing. Especially in recent years, with the rapid development of e-commerce, more and more users have begun to refer to previous users' review information to make purchase decisions. . Positive user reviews can promote the influence of commodity stores, increase sales, and bring economic benefits; on the contrary, some abnormal user reviews will affect users' online shopping experience, and criminals even hire trolls to publish false reviews to attack competitors , resulting in vicious competition. Driven by huge interests, users' abnormal comments can cause great harm to the e-commerce industry, and it...

AI Technical Summary

Problems solved by technology

However, only based on the deep neural network model to detect user comments, there are certain defects, for example, a single neural network model has its own shortcomings, the comment text has a certain hierarchy, different words form sentences, different To form a document, for different users and different products, different words and sentences in the review text contribute differently to semantic understanding, which makes it difficult for a single attention mechanism to capture global semantic information

Images