A method for automatic classification of social network user comment data

Abstract

The invention discloses a method for automatically classifying comment data of social network users. After collecting comment data in a network social platform, a set of text feature terms containing keywords is generated after preprocessing; Establish a mapping between the term set and the pre-generated category feature keywords, and judge whether the terms in the text feature term set appear in the category feature keyword library in the category, so as to calculate each of the text feature term sets The feature value of the term and its category feature keyword library, and then generate the feature vector, and use the pre-trained classification model to classify, so as to complete the automatic classification of the comment data; it has real-time performance, good classification effect, and contains more valid information.

Description

technical field [0001] The invention belongs to the technical field of natural language processing, and more specifically relates to a method for automatically classifying social network user comment data. Background technique [0002] With the widespread development of computers, especially smart phones, in our country, the Internet has become the main way for most people to obtain various information. As one of the information Internet products, Zhihu and Weibo provide a network social platform for users to use and participate in. People can obtain all kinds of information on social networking platforms, and participate in different ways such as following, commenting, and liking. With years of development, Zhihu and Weibo have repeatedly proved that they have the ability to quickly ferment and disseminate social, entertainment, news and other information, and then reach the general attention of the public. Among them, users often express their personal views or remarks i...

AI Technical Summary

Problems solved by technology

This method of purely relying on manpower to achieve the goal is a waste of user time and does not guarantee good results

[0005] At present, there is a lack of practical and effective methods and systems for automatically classifying and analyzing Chinese comment data in online social platforms. Therefore, a Chinese-based automatic text classification technology-based automatic classification and analysis method and system for online social event user comment data is found. It is an urgent task to reduce the workload of data processing personnel to the greatest extent and improve the quality of products and services

Images