Author : Gaurav Kanojia 1
Date of Publication :23rd June 2023
Abstract: It is essential, in a variety of applications that are based on the real world, such as data integration, analysis of social media, and crowdsourcing, to single out the most credible sources of information from among a group of sources that might possibly be unreliable. Truth discovery algorithms, which try to estimate the real values of a group of items by aggregating the contradictory reports supplied by multiple sources, have been offered as a solution to this issue. These algorithms are currently under development. Existing truth discovery techniques, on the other hand, often have problems with scalability and resilience, particularly when working with datasets that are of a big size and include a lot of noise. In this article, we present a novel technique that we term Scalable and Robust Truth Discovery with Stop Words and Synonyms (SRTD1); it takes use of stop words and synonyms to make the truth discovery process more accurate and efficient. In addition, we include SRTD1 with the Naive Bayes classifier in order to further improve the algorithm's already impressive level of resilience.
Reference :