| In recent years, the Internet network traffic shows a tendency of rapid growth, the economic loss of the society is becoming more and more huge because of cyber-crime is increasingly rampant,. Such as "the original paper files into","city management","motor vehicle information","hotel management information" and "on the personnel information" and so on development and application of the system, and make the public security system of the diversity of information retrieval, based on the results of the different source of reconnaissance sometimes have conflicts. These will be led to the decrease of the accuracy and rationality of reconnaissance, difficult to meet the needs of the increasingly complex network crime investigation. For the daily work of the public security system and the criminal detection, the above information for intelligent decision provides the necessary decision basis. The article first proposed the network crime of knowledge management framework. The management framework through integrated management system based on ontology is the real-time monitoring, forecasting warning, decision support and implementation of decision functions together, in order to better meet the requirements of network crime from management.Under this background, this paper put forward the algorithm of information fusion technology, which is studied and applied to the network crime detection control. For the problem that D-S evidence theory can not synthesis of highly conflict evidence, This paper focuses on the evidence sources, we proposes a new D-S improved algorithm based on the weight of evidence on the basis of the majority decide decision thoughts,and we turn to fuzzy inference system to realize the basic probability assignment. Examples of simulation show that this method can efficiently fuses conflict evidence, and compared with other improved algorithm, convergence speed is faster, more ideal synthesis results. In this paper, the main research results are as follows:(1) In view of the problem that traditional algorithm of D-S evidence theory is hard to accurately obtain the results in the case of some highly conflict evidence, we put forward the method based on discount evidence rule changes. Through introducing the concept of absolute and relative conflicts, absolute conflict represent two evidence elements intersection is empty, and the relatively conflict reaction represent the average distance of evidence, the use of the individual or the mean intensity of conflict evidence the concept of distance are cannot be effectively characterization between the degree of evidence. This paper is the product of both the geometric mean to represent the conflict between the evidence, and then use geometric mean of reciprocal to give the conflict evidence different weight, finally reuse D-S evidence combination rules of fusion, through the numerical examples validate the rationality and validity of the method.(2) According to the problem that the evidence theory algorithm is difficult to obtain the basic probability assignment function, this paper adopted the method of fuzzy inference system to obtain the basic probability assignment. Firstly carrying on reduction of the electronic data in network crime,extracting of fuzzy rules and fuzzy reasoning machine, and its output normalized probability assignment to evidence is obtained.(3) based on the JDL model and combining with the characteristics of network crime to design a three-stage multi-source information fusion decision system; Data level mainly considering the cost and time constraints, graphs distributed data mining was applied to the source data feature extraction and normalization; Character class to blur operation of property variable, fuzzy subset, determine the input and output membership functions, generate fuzzy rules, establish fuzzy inference system; Decision level by using improved D-S evidence theory to the local decision fusion reasoning results; Finally realize the accurate judgment of cyber-crime. |
PDF Full Text Request