Check_square at CheckThat! 2020: Claim Detection in Social Media via Fusion of Transformer and Syntactic Features

In this digital age of news consumption, a news reader has the ability to\nreact, express and share opinions with others in a highly interactive and fast\nmanner. As a consequence, fake news has made its way into our daily life\nbecause of very limited capacity to verify news on the Internet by large\ncompanies as well as individuals. In this paper, we focus on solving two\nproblems which are part of the fact-checking ecosystem that can help to\nautomate fact-checking of claims in an ever increasing stream of content on\nsocial media. For the first problem, claim check-worthiness prediction, we\nexplore the fusion of syntactic features and deep transformer Bidirectional\nEncoder Representations from Transformers (BERT) embeddings, to classify\ncheck-worthiness of a tweet, i.e. whether it includes a claim or not. We\nconduct a detailed feature analysis and present our best performing models for\nEnglish and Arabic tweets. For the second problem, claim retrieval, we explore\nthe pre-trained embeddings from a Siamese network transformer model\n(sentence-transformers) specifically trained for semantic textual similarity,\nand perform KD-search to retrieve verified claims with respect to a query\ntweet.\n