Analysis Of The Characteristics Of Drug-related Persons

Under the background of information technology,drug crimes have shown new characteristics in terms of criminal space and means of committing crimes.The characteristics across time and space of drug crimes and the inadequacy of information-based investigation methods have brought challenges to the detection of drug-related cases.A comprehensive analysis of the evolutionary law of development process of drug-related persons,full mining of multi-dimensional characteristics and related patterns from historical cases,and scientific verification of the spatial temporal regulations of drug-related persons and drug users have become a key link in effective prevention and response to drug crimes.Existing drug-related persons evolution analysis methods are single,characteristic mining methods are subjective,and temporal and spatial regularity verification models are lacking.The paper guided by the actual needs of the public security department for drug crime prevention,based on historical case analysis,combined with model construction,knowledge extraction,correlation analysis and other methods to extract the characteristic indicators of drug-related persons in drug-related behaviors,perform in-depth analysis of the characteristics of drugrelated persons in multiple dimensions,use the knowledge graph method for visualize the internal relationship,and verify the spatial temporal regulations through parallel systems.The relevant work and innovation points are as follows:First of all,based on the multi-Agent method,the development process model of drugrelated persons was constructed and simulated.Existing methods are difficult to accurately analyze the evolutionary laws of drug-related persons.By dividing individuals into three categories: ordinary people,drug users and drug-related persons,the paper analyzed four factors that affect individual behavior and three rules for the development of drug-related persons.The development process model of drug-related persons was established,and the model was verified by simulation.Secondly,the knowledge extraction and visual analysis of the multi-dimensional characteristics and interrelationships of drug-related persons were carried out.Based on the Bidirectional Long Short-Term Memory,Conditional Random Field and other methods,the paper constructed an entity extraction model of drug-related persons.The paper used dependency syntax and semantic analysis methods to match relationship rules,extracts related persons,locations,drug-related terms,and their relationships,and visualized the extracted entities and relationships through the Neo4 j graph database.Thirdly,efficient frequent pattern mining method was proposed,which analyzed the main characteristics and internal connections of drug-related persons from multiple angles.Based on the Apriori and FP-growth algorithms,the paper innovatively mined frequent items and association rules on the characteristics of drug-related persons,and provided reference and decision-making for investigative agencies.Finally,Poisson distribution and parallel system were combined to construct a parallel system of drug-related behaviors.Through the process of constructing an artificial society and computational experiments,the paper analyzed the temporal and spatial relationship between drug-related persons and drug users,and verified the temporal and spatial characteristics of drug transactions.