CALL FOR PAPERS : December-2022

Submission Last Date
Submit Manuscript Online





Open Access

Creative Commons License

Questcheck: Validating Unverified Texts from Social Media using Machine Reading Comprehension


Yash Muchhala , Sardar Patel Institute of Technology; Sai Nimkar, Sardar Patel Institute of Technology; Atul Kamble, Sardar Patel Institute of Technology; Dr. Radha Shankarmani, Sardar Patel Institute of Technology


Stateless, Natural Language Processing, Machine Reading Comprehension, Semantic Similarity


Social Media has become a tool for mass communication. However, some pieces of information that float around in the different mediums are not backed by facts and lack credibility. The spread of such unverified information should be curbed as it has wide-spread results in spreading misinformation that has the potential to directly affect the opinions of people consuming such information. A lot of interest in research for classifying such unverified news has been seen in the last few years. Most of the aforementioned research directly dictates that the data sources be centralized, this not only is computationally resource intensive but also requires constant effort to make sure all verified sources of truth are up to date. Therefore, in order to efficiently classify unverified news, we present a stateless technique QuestCheck, which works on the principle of Machine Reading Comprehension by aggregating the questions generated from the unverified information and pooling for answers from verified news sources on the same topic. Our validation algorithm uses a custom metric to check whether the answers obtained from both sources, i.e. verified and unverified, have a semantic similarity. Ideal scenarios for this would be validating factual information pieces, statements by massively followed personalities, and official government policies amongst other things. We present a comprehensive explanation about QuestCheck, demonstrate the working with concrete real-world examples backed by data analysis along with a detailed evaluation metric using Natural Language Processing to discuss the effect of our technique on curbing the spread of misinformation.

Other Details:

Manuscript Id :IJSTEV6I10008
Published in :Volume : 6, Issue : 10
Publication Date: 01/05/2020
Page(s): 13-17
Download Article




Submit Article

Dear Authors, You can submit your article to our journal at the following link: http://www.ijste.org/Submit

Impact Factor

The Impact Factor of our Journal is 4.753 (Year - 2016)
3.905 (Year - 2015) 2.895(Year -2014)

Click Here

Submit Payment Online

Dear Authors, Now you can submit the payment receipt to our journal online at the following link: index.php?p=Payment




Computer Science Directory. We are listed under Computer Research Institutes category

Share on Social media

Home | Privacy Policy | Terms & Conditions | Refund Policy | Feedback | Contact Us
Copyright © 2014 ijste.org All rights reserved