Crime reconstruction method and device based on Bayesian network reasoning model

Abstract

The present invention discloses a crime reconstruction method and device based on a Bayesian network reasoning model. A collection of physical evidence traces that may be produced in criminal behavior; each physical evidence trace element is used as a Bayesian network node to construct the evidence layer of the Bayesian network; the criminal facts in typical criminal behaviors are extracted to generate a hypothesis set and a behavior set, and the hypothesis set and The elements of the behavior set are used as the Bayesian network nodes to construct the hypothesis layer and behavior layer of the Bayesian network; the mother Bayesian network is constructed according to the connection relationship of the Bayesian network nodes and the network structure parameters, and the mother Bayesian network is adjusted by the growth algorithm. The network is processed to generate an instantiated Bayesian network, and crime reconstruction is performed through the instantiated Bayesian network. The model established by this method is universal and can reconstruct crimes in any of a class of cases.

Description

technical field [0001] The invention relates to the fields of criminal science and crime reconstruction technology, in particular to a crime reconstruction method and device based on a Bayesian network inference model. Background technique [0002] In a criminal case, each criminal act will produce corresponding traces of physical evidence. According to the traces of physical evidence left at the crime scene, the public security officers can analyze the crime facts such as the perpetrator, the motive for the crime, the tools for committing the crime, the time of the crime, and the location of the crime. In the field of criminal science, "criminal reconstruction" is used to refer to the above process, which can be defined as the reconstruction of the course of a crime based on the traces of physical evidence at the crime scene. The results of crime reconstruction can be divided into two categories according to different granularity: coarse-grained crime facts (such as who th...

AI Technical Summary

Problems solved by technology

[0004] (1) The generality of the model is insufficient: for each case, a corresponding reasoning model needs to be constructed;

[0005] (2) The degree of automation of the model is insufficient: each model needs to be constructed manually, and the machine cannot automatically model, which is time-consuming and laborious;

[0006] (3) The results given by the model are not comprehensive: the model does not consider both fine-grained criminal behavior and coarse-grained criminal facts

Images