System and Method for Truth Discovery in social media Big Data

Abstract

Within the span of enormous information and the coming of numerous advancements in the communication technologies, at every tick of the clock, enormous sums of information is produced from different sources. One such source of data generation is social media. However, such data carries much of the noisy, uncertain, and untrustworthy data. In this way, finding independable information from loud information is one of the characteristic challenges of huge information focusing on the esteem characteristic of enormous information. Therefore, in this article, an attempt is made to target a few challenges arriving from “misinformation spread”, “data sparsity” or the “long-tail wonder” in the domain of social media data analytics. The study uses an instance from the Online Social Network (OSN) datasets to develop scalable to wide-range social sensing by consolidating Scalable Robust Trust Discovery (SRTD) plots to address the mentioned challenges utilizing the distributed parallel computing framework. The dataset picked for investigation includes 128,483 tweets which incorporates 20% deception, 80% retweets bringing about 0.05 milliseconds utilizing Spark parallel processing.