On scalable and robust truth discovery in big data social media sensing applications

Abstract

Recognizing the date responsible for continuing the set for an unventilated spring variety of online media has been an essential mission within the Big Data period. This task, whose maiden name is precision detection, aims to recognize the uniformity of spring and therefore the honesty of declaring that they are built without being communicative a priori. In this crack, we recognize three core face up to that have not been well deal with in the literature on the innovation of the truth in progress. The first is the "propagation of the party line", where the most important variety of resource is contributing to false maintain, which makes it easier said than done to identify true claims. Challenge is the "data shortage" or the "long tail phenomenon", where most sources only provide a small number of claims, which provides insufficient evidence to determine the reliability of those sources.