False comment identification method based on rolling type cooperative training

Abstract

The invention discloses a false comment identification method based on rolling type cooperative training, and relates to the technical field of false comment identification. The method comprises the steps: obtaining a comment text, and determining the content features of the comment text according to the comment text; obtaining reviewer information corresponding to the comment text, and determining reviewer behavior characteristics according to the reviewer information; performing false comment identification according to the comment text content characteristics and the reviewer behavior characteristics; outputting false comment recognition results. According to the false comment recognition method, unlabeled samples are effectively utilized to assist model learning, multiple features suchas emotion and text representation are fused at the same time, the recognition performance of the model is improved through multi-feature fusion collaborative training, and the accuracy is improved by 3.5% compared with a traditional false comment recognition method.

Description

technical field [0001] The invention relates to the technical field of false comment identification, in particular to a method for identifying false comments based on rolling collaborative training. Background technique [0002] For online shopping, due to the inconsistency between the online product information and the products received by consumers offline, consumers will read a large number of reviews of target products to assist in judgment. Therefore, commodity reviews affect consumers' purchasing behavior, and also affect the interests of merchants. Positive reviews attract more potential users, while negative reviews drive away potential users. In order to obtain higher profits, unscrupulous merchants usually hire professional writers to write false positive reviews for their own products to attract potential users; to write false negative reviews for competitors to suppress competitors. For example, “Some merchants use the method of “rebate for good reviews” to obt...

AI Technical Summary

Problems solved by technology

[0005] (1) The classification method under the fully supervised framework is mainly used. The fully supervised learning method requires a large amount of labeled corpus as a training set. The lack of standard data sets and the time-consuming and laborious problems of manual labeling have brought great challenges to the research under the fully supervised framework. Earth limitations

[0006] (2) Scholars try to use unsupervised learning methods. Unsupervised learning methods solve the problem of missing labeled datasets, but the recognition accuracy is generally low.

[0007] (3) Semi-supervised learning balances the main problems of fully supervised learning and unsupervised learning. However, in the current research on identifying false reviews with semi-supervised methods, simple feature construction is only based on shallow review features. model, ignoring the impact of differences in the combination of different features and classification models on the results

Images