The invention discloses a financial news tendency analysis method based on LSTMLSTM-based financial news tendency analysis method

Abstract

The invention relates to an LSTM-based financial news tendency analysis method. The method comprises the steps of performing company name identification based on Baidu encyclopedia query and company name and company code mapping; C; comparing the similarity between sentences and titles by using a doc2vec model, and extracting key sentence groups by comprehensively considering sentence positions, domain verbs and company name information; A; and using word2vec and TFIDF to represent sentences, and using an LSTM model to classify the key sentence group. According to the invention, Baidu encyclopedia query is added as a factor of identification in the company name identification method; t; the effect is better; t; the expansibility is better; t; the problem that the names of non-companies aremisjudged due to too many products is solved; A; according to the method, t, the key sentence group is extracted and introduced into the doc2vec model, t, the similarity calculation accuracy is improved, when sentences are represented, t, the Word2vec is used for training the text, meanwhile, t, the TFIDF method is introduced, t, text context information and the importance degree of words in thetext are fully utilized, and a very good effect can be achieved.

Description

technical field [0001] The invention belongs to the technical field of text processing, and in particular relates to an LSTM-based financial news tendency analysis method. Background technique [0002] The negative tendency of financial news will damage the reputation of banks and financial institutions. At the same time, it will also affect the decision-making of investors to a certain extent. Especially in today's information age, a small matter may be infinitely fermented and spread in a short period of time. For countless audiences, a news may cause a company to emerge into crisis or even face bankruptcy. For example: in 2016, a news titled LeEco’s capital chain rupture was spread on the Internet, causing LeEco’s share price to shrink sharply, closing down 7.49% on the day. Therefore, it is of great significance to perceive risks in advance for enterprises. With the development of the Internet, more and more people can easily get access to financial news. It is difficul...

AI Technical Summary

Problems solved by technology

However, the full name of the company rarely appears in the text of financial news, which greatly limits the scope of application of this method

(2) Manually summarize the structural characteristics of the company name Based on the context information and six knowledge bases, two scans are performed to identify the company name. This method focuses on identifying the full name of the company, and the accuracy and recall rates for identifying abbreviations are both high. relatively low

It relies heavily on full company names and does not recognize new company names

(4) Use Internet search and listed company name information to identify. The company name dictionary of this method does not consider the company stock code and the information queried by Baidu Encyclopedia. Therefore, there are still some famous companies that are introduced in the results returned by Baidu search due to their large number of products. Many products are not recognized as company names

[0005] It can be seen that the existing company name recognition methods in the prior art still have relatively low recognition accuracy for company abbreviations, and the rules and knowledge bases in some methods are too complicated, which seriously affects the application of the method

Images