Method for identifying substance features of customer reviews

Abstract

The invention discloses a method for identifying substance features of customer reviews, which is mainly used for solving the problem of poor accuracy of the conventional automatic substance feature extraction technology on Chinese customer reviews. The method comprises the following implementation steps of: sampling an appropriate quantity of reviews serving as a training set, and performing secondary part-of-speech tagging on the training set; excavating a frequent item set based on a Classification Based on Association (CBA) method; and correcting centralized substance features in the frequent item set according to a dependence relationship to form a candidate substance feature set, performing reliability judgment, removing unreliable substance features, combining synonymic substance features through lexical meaning similarity computation to form a substance feature adjacency set, and screening valuable customer reviews from the substance feature adjacency set to prevent customers from reading a large amount of invaluable information and realize specific lookup and reading of certain substance features for the customers. The method is suitable for identifying the substance features of Chinese network reviews, and has the advantages of high accuracy, portability, simple structure and convenience in implementing.

Description

technical field [0001] The invention belongs to the field of computer technology, relates to data mining and natural language processing, and can be used to identify the main entity features of user comments. Background technique [0002] With the rapid development of information technology, the Internet, as the fourth type of digital media, has become the largest communication media in the world today. People use the Internet for distance learning, use the Internet to understand current social events, and also express different views on a certain point of view in a certain topic through the Internet, such as feedback on network teaching suggestions or opinions on current events. [0003] More and more online public opinion hotspots show that the Internet has become an important channel for Chinese netizens to participate in society. According to data released by the China Internet Network Information Center a few days ago, 56% of Internet users often post comments on the I...

AI Technical Summary

Problems solved by technology

The method of manually defining entity characteristics requires the participation of experts in this field for entities in each field, so it is not portable

At the same time, manual definition of entity features does not have the function of real-time update with the addition of new product features

Automatic recognition mainly uses natural language technologies such as part-of-speech tagging, syntactic analysis, and text mode to analyze the sentences in entity comments, and automatically discover entity features from them. It has good versatility and portability, but the disadvantage is that the accuracy rate is not high. Moreover, most of the current automatic extraction technologies are researched on English user comments, and there is still a lack of methods for identifying the entity characteristics of Chinese online comments.

Images