Enterprise entity name analysis and identification system

Abstract

The invention relates to the field of natural language processing, in particular to an enterprise entity name analysis and identification system. The system comprises a bidirectional recurrent neural network module; the system uses an enterprise name labeling training sample stored in an existing enterprise name database to train a bidirectional recurrent neural network; the bidirectional recurrent neural network identifies enterprise names in to-be-processed texts and extracts names which do not belong to existing enterprise names; the system enables a classification and judgment result of to-be-identified natural language sequences to depend on context information and enables the preparation rate of extraction and judgment to be higher by performing characteristic automatic learning on basic elements of the texts, such as characters, words, punctuation marks and the like, and applying the bidirectional RNN (Recurrent Neural Network); and the system discovers new enterprise entity names through existing data characteristics and has important application values in the field of big data analysis, particularly in the field of data analysis taking enterprises as analysis main bodies.

Description

technical field [0001] The invention relates to the field of natural language processing, in particular to an enterprise entity name analysis and recognition system. 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 of these Internet industries is inseparable from the analysis of a large amount of information and data, and the value of information analysis lies in its accuracy and sensitivity. Sensitive analysis requires timely and rapid discovery of new information; however, most of the data obtained directly from web pages are very Structured, in order to use these data, data cleaning has become the place where companies spend the most time and energy. In data cleaning, the extraction of specific information,...

AI Technical Summary

Problems solved by technology

To use conditional random fields, you first need to design and construct feature templates based on the characteristics of the entity to be recognized. Feature templates include first-order words or multi-order phrases with a specified window size context, word prefixes, suffixes, part-of-speech tags and other status features; feature templates The construction is very time-consuming and labor-intensive, and the recognition results are highly dependent on the feature templates, while the manually set feature templates are often only based on the characteristics of some samples, and have poor versatility; The use is also independent of each other. The prediction cannot rely on longer historical state information, nor can it use longer future information feedback to correct possible historical errors; the prediction process is time-consuming and laborious, and the prediction results are difficult to achieve global optimality

Images