Research And Application Of Data Analysis For Android Forensics

In recent years, smart phones and other mobile devices based on the Android operating system change from a simple communication tool into a comprehensive platform for information processing and services. With the high-tech crimes and civil disputes increased, criminal means increasingly become digitized and diversity, and demands of electronic evidence involved is gradually increasing.There are three stages in Digital forensic: data acquisition and data recovery, data analysis, output and display the results. As a key stage of digital forensic, how to extract significant useful information is a concern problem in the face of a large number of electronic documents. By studying the field of information technology is widely used in text classification, it believes it can effectively deal with data analysis problems.This paper focus on research and application of data analysis for Android forensics. Firstly,this paper talks about digital forensics technology, makes a brief introduction to the data recovery and acquisition and data analysis in recently years and introduces the different and connection between computer forensics and cellphone forensics. Then, this paper shows the framework and file structure of Android system and show me what is the important point about Android forensics.This article introduces the process of text classification, commonly used feature selection algorithm, three classification algorithms and criterion of performance evaluation. Then, we propose a new tf*idf algorithm—improved tf*idf. Experimental results show that: using new tf*idf algorithm is better than tranditional tf*idf algorithm. Expect that, this paper put forward the characteristics expansion algorithm based on the wikipedia. The paper finally designs the data analysis subsystem based on Android, used for digital forensics. The subsystem has been improved gradually with repeated test, and it has practicability and important significance for judicial department and public security department.The automation of smartphone forensics is the most important step in the whole forensics process, which not only need to reduce the forensics work, but also need to ensure the fullly and effectively access to data. Use text classification algorithm to achieve smart phone forensics automated has a positive meaning.