Research On Object Classification And Sensitive Topic Analysis Method Based On Forensic Data

The advent of the third information technology revolution has led to the rapid development of Internet technology.With the improvement of people's living standards,mobile smart terminals represented by smartphones have been rapidly popularized.Intelligent mobile terminals have greatly improved people's convenience and brought many conveniences to humans.But there are always criminals who use this convenience to carry out illegal activities,such as Internet fraud,online violence,and pornographic transmission.In order to deal with the chaos caused by the Internet,mobile forensics technology has been used to combat criminal behaviors and provide powerful weapons for criminal investigators.As the mobile forensics industry started late,there is still some room for development.At present,there are digital forensics companies dedicated to the analysis and mining of forensic data.Due to the increasing number of information types,it is difficult to mine the potential information between forensic data in all aspects.There is still a large amount of valuable data that has not been mined,so it is still necessary to improve forensics.Data mining ability.This article is devoted to improving the data mining ability of forensic data.First,it introduces the research background and significance of mobile digital forensics and data mining in detail,analyzes the current status and development trends of digital forensics at home and abroad,and introduces forensic data.and data mining in detail.Related basic knowledge and implementation technology,including the concept of forensic data,word2vec concept,semantic similarity concept,LDA algorithm and cloud model.Aiming at the problems of tag cloud merging and multi-tag output in the classification of communication objects by combining forensic data and cloud model text classification methods,this paper proposes a multi-classification method of forensic data objects based on clustering cloud model,and verifies the method through experiments The feasibility of the method effectively mines the relationship between the communication object and the head of the household.For the problem of the inconsistency of the number of topics in each piece of data in the forensic data,this paper proposes a method based on the characteristics of forensic data based on the traditional LDA topic analysis method LDA's forensic data object sensitive subject analysis method can easily and effectively mine the sensitive subject content of forensic data.The research on object classification and sensitive analysis methods based on forensic data proposed in this paper mainly mines two aspects of forensic data,namely,mining the relationship between forensic data objects and household heads and mining potentially sensitive topics in chat records of various objects.Not only is it possible to mine the potential homeowner relationship network of forensic data,but it also helps criminal investigators to more quickly understand the interactive content of suspected targets,which can improve the accuracy and efficiency of digital forensics.