An Adaptive Sliding Window Algorithm for Mining Frequent Itemsets in Computer Forensics

摘要

Data mining technology is widely utilized in the field of computer criminal forensics. The research of data mining technologies, such as frequent itemset mining, clustering etc., can effectively improve the efficiency of computer forensics. However, data items of streaming data are dynamically changed along with time, which triggers off some new challenges to computer forensics. To address this issue, this paper presents an adaptive sliding window based strategy for mining the main frequent itemsets on streaming data. The key idea is to dynamically adjust the size of sliding window by exploiting the time-varying feature of streaming data, in order to satisfy the concept change that occurs in the streaming data. The experimental results show that compared with the previous work, the proposed algorithm can superiorly adapt to the time-varying feature of streaming data, and dramatically enhance time performance by reducing the data size for mining.