Method for analyzing tendentiousness of financial news

Abstract

The invention relates to a method for analyzing tendentiousness of financial news. The method comprises: identifying a company name, extracting key sentence groups and classifying the key sentence groups by using an LSTM model. The invention provides a method for analyzing tendentiousness of financial news, and the company name is identified by using methods of company name abbreviation dictionaryand encyclopedia query, and the method has good effect and expansibility. A key sentence group extraction method based on deep learning framework doc2vec text similarity matched with comprehensive feature attributes is used, and the method has good extraction effect, high accuracy and recall rate, high accuracy rate of text tendency determination, and good effect, and can preferably meet needs ofpractical application.

Description

Technical field [0001] The invention belongs to the technical field of text analysis, and specifically relates to a method for analyzing financial news tendency. Background technique [0002] Current text sentiment orientation analysis methods include the following: 1) Use attributes such as sentiment, location, and keywords as factors for extracting key sentences, and then perform supervised and semi-supervised sentiment classification of key sentence groups. This method takes the key The accuracy of the sentence is not high; 2) Use emotional vocabulary training text containing negative vocabulary, tendency vocabulary, and degree vocabulary for feature expansion. This method does not consider the context and does not work well. There are relatively few studies on the classification of financial news texts at home and abroad, mainly using vocabulary and semantic rules to extract the semantic annotation information of events in the news text, and using this information for the fea...

AI Technical Summary

Problems solved by technology

[0002] The current text sentiment analysis methods include the following: 1) Use attributes such as emotion, location, and keywords as factors for extracting key sentences, and then carry out supervised and semi-supervised emotional classification of key sentence groups. The accuracy rate of the sentence is not high; 2) Use the emotional vocabulary training text containing negative vocabulary, tendency vocabulary, and degree vocabulary to perform feature expansion. This method does not consider the context, and the effect is not good

There are relatively few studies on targeted financial news text classification at home and abroad, mainly using lexical and semantic rules to extract event semantic annotation information in news text, and using this information for machi...

Images