Dictionary-based multi-dimensional emotion analysis method

Abstract

The invention relates to the field of natural language processing, in particular to a dictionary-based multi-dimensional emotion analysis method. The method comprises the following steps of randomly drawing a certain number of texts from related documents of a to-be-analyzed target group object to form an initial text set and clustering to form N subject text sets on the basis of automatically extracting text keywords; constructing M domains according to the analysis needs, selecting seed words of each domain, comparing the relationships between the seed words of the domains and the subject text sets and taking the closest relationship as a source text set extended by a corresponding domain dictionary; constructing each domain dictionary in the source text set according to the expansion of the seed words; and applying the constructed domain dictionaries to carry out feature extraction on the related documents of the target object and combining with a correlated model algorithm to give a judging result of emotion analysis of the corresponding domain. According to the method, a plurality of targeted domain dictionaries are quickly constructed as needed, the multi-dimensional comprehensive emotion analysis of the specific object is realized, the visual angle is more comprehensive and the result is more accurate.

Description

technical field [0001] The invention relates to the field of natural language processing, in particular to a dictionary-based multi-dimensional sentiment analysis method. Background technique [0002] With the rapid development of the Internet, a large amount of public web data has been generated, which has also spurred various emerging industries based on big data technology, such as Internet medical care, Internet education, corporate or personal credit investigation, etc. The rise and prosperity of these Internet industries are inseparable from the analysis of a large amount of data information. Natural language processing plays an important role in big data analysis. In the face of massive network text resources, we can automatically and intelligently extract the information we care about by using natural language processing analysis methods, or judge the text or text publishers. This kind of emotional tendency has crucial practical application significance in both publ...

AI Technical Summary

Problems solved by technology

The sentiment analysis method based on the sentiment dictionary is targeted analysis and excavation. The dictionaries adopted in different fields are also very different. At present, the existing field dictionaries lack the applicability to specific problems and are not very targeted.

When analyzing specific fields or specific topics, using existing large and broad domain dictionaries cannot achieve good analysis results

[0005] In addition, there are more and more dimensions required for sentiment analysis at this stage, and the analysis direction for the same target group is becoming more and more detailed. Different dictionaries need to be constructed for different analysis dimensions to perform corresponding feature extraction. Among the existing methods The dictionaries of different analysis directions are constructed separately, and the way of manually constructing dictionaries one by one is very time-consuming and labor-intensive, which increases the difficulty of comprehensive and multi-faceted analysis of the target group, increases the working hours of the project in terms of technical implementation, and extends project delivery

Images