Financial field term recognizing method based on information entropy and term credibility

Abstract

The invention provides a financial field term recognizing method based on information entropy and term credibility. Only simple characteristics are selected, and financial terms are recognized through a CRF model; candidate terms belonging to the specific error type are screened out by setting a threshold according to an information entropy formula based on marginal probability in a recognition result, and the candidate terms are processed in a more targeted mode; words are converted into word vectors with rich semantic information when the candidate terms are filtered, and a large number of financial field terms can be obtained through filtering since a similarity calculation method and a traditional mutual information method complement each other. The too complicated characteristic selection process of an existing robot learning model can be effectively avoided, post-processing part is flexible and not limited to specific linguistic data, the recall rate can be easily increased, the term structure integrity can be improved, and the method can be used as a universal term recognizing method.

Description

technical field [0001] The invention relates to the fields of natural language processing, text mining, information processing, named entity recognition, etc., focuses on the problem of term recognition, and proposes a term recognition method in the financial field based on information entropy and term credibility. The method effectively improves the recall rate and the integrity of the term structure, and can be used as a general term recognition method. Background technique [0002] With the in-depth development of the market economy, the position of finance in social life has become increasingly prominent, and finance has become the most important strategic resource for economic operation and a powerful booster for regional economic development. Compared with other fields, terms in the financial field are replaced faster. Rapid identification of financial terms has high application value for tasks such as text mining, information extraction, and public opinion analysis in...

AI Technical Summary

Problems solved by technology

This type of model transforms the problem of term recognition into a sequence labeling problem. It usually requires manual addition of features to fit the training data, and the selection and combination of these features is time-consuming and labor-intensive. Too many features will also increase the complexity of the model, resulting in excessive Fitting phenomenon

Images