SU-NLP at CheckThat! 2021: check-worthiness of Turkish tweets

The growth in social media usage increases the spread of misinformation on these platforms. In order to prevent this disinformation spread, automated fact checking systems that identify and verify claims are needed. The first step of such systems is the identification of whether a claim is worth-checking or not. This paper describes our participation to the check-worthiness task of CLEF 2021 CheckThat! 2021 Lab for Turkish tweets. We propose an ensemble of BERT models which ranked the second best in terms of MAP score.