New word discovery method based on model

Abstract

The invention discloses a new word discovery method based on a model, which belongs to the technical field of auditing, and comprises the following steps: S1, acquiring an audit service related document; and S2, performing format conversion on the audit service related document, and taking the converted audit service related document data as input data of a subsequent labeling tool. According to the method, on the basis of audit data, continuous discovery of audit field words is realized by utilizing a new word discovery technology, the professional word bank in the audit field is preliminarily constructed in combination with a combed existing industry word bank, the word bank is subsequently audited by professionals, and the professional word bank in the audit field is finally formed, so that effective support is provided for subsequent audit data analysis, the text is subjected to preliminary 'new word' discovery, then the 'new word' discovered manually is checked, and the real professional vocabularies in the auditing field are extracted, so that the workload of purely manually extracting the professional vocabularies in the auditing field from the document can be reduced to a great extent, and the construction efficiency of an auditing professional lexicon is improved.

Description

technical field [0001] The invention relates to the technical field of auditing, in particular to a model-based new word discovery method. Background technique [0002] In recent years, with the continuous application and development of information technologies such as big data, artificial intelligence, cloud computing, Internet of Things, and mobile applications, people's lives and work have gradually changed, bringing opportunities and challenges to audit supervision. Internal audit work is facing With the profound reform of audit informatization. [0003] In the process of audit informatization, it is necessary to extract domain words. For example, the rule-based extraction method is to establish corresponding rules based on the composition structure of words and external context connections, and use pattern matching to extract domain words. This extraction method Most of the rules are made manually, and it is difficult to automatically discover the rules with a computer...

AI Technical Summary

Problems solved by technology

[0003] In the process of audit informatization, it is necessary to extract domain words. For example, the rule-based extraction method is to establish corresponding rules based on the composition structure of words and external context connections, and use pattern matching to extract domain words. This extraction method Most of the rules are made manually, and it is difficult to automatically discover the rules with a computer. Especially nowadays, Internet buzzwords are so strange that it is even more difficult to find their regularity, so it is very difficult. However, this method is not very effective in identifying individual domain words and low-frequency domain words, so there is an urgent need for a high-quality model-based new word discovery method

Images