Holographic Portrait Method Of Drug-Related Personnel Based On Big Data

In recent years,the number of drug-related crimes in China has surged.Under the background of information age,drug crimes show many new features,which bring new challenges to the detection of drug-related cases.It is an important basic work to crack down on and prevent drug-related crimes by fully excavating the characteristics and connections of drugrelated personnel from past cases and making scientific analysis.Aiming at the characteristics of modern drug-related crimes,such as grouping and concealment of drug crimes,this dissertation proposes a holographic portrait method of drug-related personnel based on big data.By collecting the criminal judgment documents of drug-related cases in the judgment document network,knowledge is extracted,integrated and stored to form a drug-related social network;By adding public security data,the database of public security anti-drug knowledge can be constructed,and the holographic portrait of complex relationship among drug-related personnel can be realized,and the behaviors of drug-related personnel can be analyzed and predicted.The main innovations and works of this dissertation are as follows:1.A collection method of drug-related documents based on improved K-means algorithm is proposed.Build a real-time document collection and classification system to ensure the dynamic collection of drug-related personnel data.By extracting the elements of sentencing,it is found that the sentencing results are influenced by the quantity and type of drugs,confession and punishment,recidivism and other factors.Aiming at the phenomenon of multiple people committing multiple crimes,this dissertation puts forward a correlation analysis model between drug-related behaviors and drug-related personnel,and extracts the information of drug-related personnel as the subject of drug-related behaviors.2.Construct static holographic portraits and dynamic holographic portraits of drug-related personnel.Firstly,based on GRU model,learning features are connected to CRF decoding layer to build entity extraction model,and dependency syntax analysis and other methods are used to match relationship rules to complete entity extraction and relationship extraction of drug-related personnel,and KGCloud cloud platform is used to display static holographic portraits of drugrelated personnel.Secondly,based on BERT+Bi GRU+CRF method,Chinese address elements are analyzed,which is used as the spatial elements of dynamic holographic portraits.Finally,the time series prediction model of the number of drug-related people is constructed,and the validity of the model is verified by fitting the predicted data of infectious diseases based on Prophet model,which provides a time tag reference for the construction of dynamic holographic portraits of drug-related people.3.Build a social network of drug-related personnel to simulate the spreading process of drug-related behaviors.Based on SIR infectious disease model and Agent modeling theory,this dissertation constructs the social network of drug-related personnel,puts forward a drug-related behavior transmission model,and analyzes the spread of drug-related behaviors under different conditions.4.Design an anti-drug monitoring platform system for public security organs.Through the display of holographic portraits of drug-related personnel,help the police handling cases of public security organs to fully understand all relevant information of drug-related personnel;Make use of knowledge map to analyze and judge,and excavate and hide drug-related personnel and gangs;By analyzing the spread of drug-related behaviors in social networks,we can predict the spread trend of drug-related behaviors and help public security organs take measures to prevent drug-related crimes in advance.