Method, device and electronic equipment for triplet extraction of safety report text

Abstract

The invention discloses a method for extracting triples of safety report text, including: obtaining safety report text; performing sentence segmentation processing on the safety report text; Perform word segmentation processing and part-of-speech tagging, and output the results of word segmentation and part-of-speech tagging; perform syntactic analysis on the results of word segmentation and part-of-speech tagging based on the dependency syntax analysis method to obtain the grammatical components and the relationship between the grammatical components; perform semantic role labeling to obtain the given Argument of the predicate; outputs a triple of subject-predicate-object form. Corresponding devices, electronic equipment and computer-readable storage media are also disclosed. The text of the security report is analyzed by means of syntactic analysis, triples are extracted, and proper nouns in the security field are added in the process of word segmentation, so that the model can be better recognized. The location and length of the subject can better express the meaning of the subject, verb and object in the triple to meet the needs of accurate intelligence analysis and screening.

Description

technical field [0001] The invention relates to the technical field of natural language recognition and analysis processing, in particular to a method, device and electronic equipment for triplet extraction of safety report text. Background technique [0002] Threat intelligence is security information that has been researched and judged. Threat intelligence can be used to support decision-making or security analysis. Security information of unknown origin and authenticity will affect the correctness of decision-making and the accuracy of analysis results. Information research and judgment is divided into two aspects, the source of information and the information itself. It can ensure that the source of information is reliable, but the existing technology usually uses manual methods to screen and summarize threat intelligence from massive security reports, which is very time-consuming and laborious, and the accuracy of screening is biased due to the inability to always focus...

AI Technical Summary

Problems solved by technology

It can ensure that the source of information is reliable, but the existing technology usually uses manual methods to screen and summarize threat intelligence from massive security reports, which is very time-consuming and laborious, and the accuracy of screening is biased due to the inability to always focus on the information itself. Low, that is, it is impossible to ensure that the key information can be extracted manually. Therefore, it is necessary to optimize the method and method of text extraction in the security field, which has become a new development direction.

Images