Recognizing Fabricated Statements by Prominent Figures
Main Article Content
Abstract
This paper presents a method aimed at detecting fraudulent explanations provided by public figures using fabricated data. Several strategies were implemented into a computer program system and tested on a dataset of explanations. The best achieved outcome in binary classification (valid or false explanation) is 92% accuracy with the Xtreme Angle Boosting algorithm. Further enhancements are discussed within the article.
Article Details
Issue
Section
Articles