Security analysis and automatic evaluation method based on index threshold value and semantic analysis

Abstract

The invention discloses a safety analysis and automatic evaluation method based on an index threshold value and semantic analysis. The method comprises the following steps: firstly, extracting a semantic relationship from an unstructured security analysis report, a regulation standard and other security analysis original corpus, and constructing a normalized security knowledge base; meanwhile, a safety index entity set is constructed according to the dimensions of activities, items and the like, and the safety index set is classified and divided in a layered mode; based on a semantic similarity calculation method, safety index entities and threshold attributes are identified and extracted from corpora such as regulatory standards, technical requirements and judgment bases, and the safety index entities and monitoring data attributes can be identified and extracted from a safety analysis report; and according to a semantic analysis result, automatic benchmarking and judgment can be carried out on a safety index satisfaction condition, and an expert experience method can be combined to carry out re-checking on an index initial judgment result. According to the method, the requirements of users on rapid, accurate and intelligent safety analysis and automatic evaluation of different activities and items are met, and the safety evaluation efficiency and accuracy can be effectively improved.

Description

technical field [0001] The invention belongs to the field of safety supervision, and in particular relates to a safety analysis and automatic evaluation method based on index thresholds and semantic analysis, which is suitable for safety evaluation of certain types of task activities or task matters. Background technique [0002] For the task activities or tasks in the field of safety supervision, in the safety assessment work that has been carried out, the method of document review is usually used, and the review experts compare the relevant safety laws, regulations and technical standards, and combine the expert experience to find safety issues, and through on-site witnesses With on-site measurement and other methods to further verify related problems, with more and more input of safety assessment tasks and the accumulation and update of safety assessment-related materials and data resources, how to establish more scientific and effective analysis and evaluation capabilitie...

AI Technical Summary

Problems solved by technology

Existing document review methods have increasingly obvious deficiencies such as excessive reliance on expert experience, insufficient information mining, low data utilization, and insufficient automation of analysis methods

Images