Enhancing trust accuracy among online social network users utilizing data text mining techniques in apache spark

摘要

The number of users and amount of data transfer are increasing per each minute with the rapid growth of social network platforms on the web while the users have no certain knowledge of each other. Thus, with the overwhelming spread of the internet and such bulk of data, people find it arduous to identify valid comments. Establishing a genuine and more accurate trust becomes harder if classical processing is used especially with the presence of profitable, oriented, devious and narrow-minded comments. Various methods have been employed so far to evaluate reliable users most of which combine trust algorithms, subject classification, and comment mining methods. Researches reveal that the majority of social network users firstly take into account an overall number of public trust standards such as the number of friends, followers, followings, and likes of individuals in order to trust them. However, a malicious user could manipulate this trust by building virtual qualities. Accordingly, this study supplies a dictionary of malicious words and weighs them by combining trust standards and text mining users' tweets. It is intended to identify malicious users and analyze their behavior to proceed a more accurate trust within distributed execution in Spark environment for providing a quicker call. The results of this study show that the suggested method benefits from a high diagnostic accuracy.