Violation corpus detection method and device, computer equipment and storage medium

Abstract

The embodiment of the invention belongs to the field of artificial intelligence, is applied to the field of smart government affairs, and relates to a violation corpus detection method and device, computer equipment and a storage medium.The method comprises the following steps: generating a tagged corpus carrying a first-level tag based on a historical voice text, the first-level tag comprising a violation tag and a non-violation tag, wherein the tagged corpus carrying the violation tag carries a secondary tag; adjusting the annotation corpus to obtain a target training corpus; training an initial text classification model based on the target training corpus and the first-level tag to obtain a target text classification model, and training an initial statement classification model based on the target training corpus and the second-level tag to obtain a target statement classification model; inputting the corpus to be recognized into the target text classification model to obtain a target first-level tag; and when the target first-level tag is a violation tag, inputting the to-be-identified corpus into the target statement classification model to obtain a target second-level tag, wherein the target statement classification model is stored in a block chain. According to the method and the device, identification of violation corpora is effectively improved.

Description

Technical field[0001]The present application relates to artificial intelligence technology, and in particular, it relates to detection methods, devices, computer equipment and storage media of illegal corpus.Background technique[0002]With the continuous innovation and development of computer technology, computers have been applied to all walks of life. During the access to the customer, the seat is often easy to communicate in violation, causing the customer's experience. Therefore, the quality inspection of the corpus generated during communication is a very important part. At present, the quality inspection of the corpus often uses the cultural classification model after training to conduct a quality inspection classification, thereby determining the corpus of violations.[0003]However, due to the large amount of data of the corpus, it is often only a small part of the corpus, leading to extreme imbalances, and the corpus-classification model is not good for small probability of il...

AI Technical Summary

Problems solved by technology

[0003] However, due to the huge amount of data in the corpus, the corpus that violates the rules often only occupies a small part, resulting in extremely unbalanced training samples. The case of hard recall of probabilistic offending corpus and false recall of non-offending corpus

Images