期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：10
页码：15936
DOI：10.15680/IJIRCCE.2017.0510049
出版社：S&S Publications
摘要：In recent trends, mobile devices are very popular and the growth of those mobile devices is tremendousin nature. Due to this marvelous growth, data analysis and user preference and interest detectingon are very new andinteresting. The Proposed System helps to detect anomaly and user preferences by using Wi-Fi logs from mobiledevices. The analysis of user behavior from Wi-Fi logs are much complicated because the Wi-Fi logs contains manyauxiliary information and noises. There is an enormous challenge to perform the elimination of auxiliary information’sand performing user summary from Wi-Fi logs. In this proposal, there are 3 algorithms are used after data cleaning forenabling the user preference and anomaly detection. The Access Pattern Analysis (APA) analysis Algorithm isproposed to identify user and session are very important for identifying behavioral patterns. The time taken foridentifying user and session are considerably reduced due to the effect of APA. Improved Expectation Maximization(IEM) clustering algorithm is proposed to help in identifying very relevant similar groups. The similarities betweenuser access and their behaviors are identified using IEM. The proposed Multi Label PCA strategy is used for findingand pruning unwanted SSID and log details. And it also helps to detect every pattern from APA. The experiments showthe Proposed System can effectively find the location based user preferences and anomalies with Wi-Fi and Web logs.
