Knowledge graph-based guidance and instigation corpus detection method and related equipment thereof

Abstract

The embodiment of the invention belongs to the field of big data, is applied to the field of smart government affairs / education, and relates to a knowledge graph-based guidance and instigation corpusdetection method and related equipment, and the method comprises the steps of receiving a standard corpus data set, performing feature extraction on the standard corpus data set, and obtaining standard corpus features, wherein no guidance and instigation information exists in the standard corpus data set; constructing a first knowledge graph based on the standard corpus features; receiving to-be-detected corpora, performing named entity identification on the to-be-detected corpora to obtain to-be-detected entities, and performing deduction on each to-be-detected entity in the first knowledge graph; and when the to-be-detected entities fail in deduction, taking the to-be-detected entities failing in deduction as guidance and instigation entities, taking to-be-detected corpora correspondingto the guidance and instigation entities as guidance and instigation corpora, and outputting the guidance and instigation corpora, wherein the first knowledge graph can be stored in a blockchain. According to the method, the guidance and instigation corpus can be quickly identified, so that the guidance and instigation behavior is detected.

Description

technical field [0001] This application relates to the field of big data technology, and in particular to a detection method and related equipment for guiding and abetting corpus based on knowledge graphs. Background technique [0002] With the continuous innovation and development of computer technology, computers have been applied to all walks of life. In the process of communication between agents and customers, it is often easy to guide and abet customers. Therefore, guidance and abetment is a common violation scenario in voice quality inspection. The frequency of this violation is high, and the nature of the violation is relatively serious. An important quality inspection point in the inspection process. [0003] Traditional quality inspection algorithms are mostly based on regular matching rules, which have the limitations of relatively single coverage scenarios and overlapping generalization capabilities. At the same time, with the continuous optimization of agent s...

AI Technical Summary

Problems solved by technology

If a completely rule-based algorithm is used for detection, it will take a huge amount of manpower to collect, mark, guide and instigate illegal words, and write lengthy and complex rule logic. Computers cannot self-learn, update and optimize over time.

Images