Brainstormers_msec at SemEval-2023 Task 10: Detection of sexism related comments in social media using deep learning

Proceedings of the The 17th International Workshop on Semantic Evaluation (SemEval-2023), pages 1114–1120 July 13-14, 2023 ©2023 Association for Computational Linguistics Brainstormers_msec at SemEval-2023 Task 10: Detection of sexism related comments in social media using deep learning C.Jerin Mahibha and C.M.Swaathi and R.Jeevitha and R.Princy Martina Meenakshi Sundararajan Engineering College, Chennai [jerinmahibha, cmswaathi0306, rahijee365,princymartina2001] @gmail.com Durairaj Thenmozhi Sri Sivasubramaniya Nadar College of Engineering , Chennai theni_d@ssn.edu.in Abstract Social media is the media through which peo- ple share their thoughts and opinions. This has both its pros and cons which depends on the type of information being conveyed. If any in- formation conveyed over social media hurts or affects a person, such information can be re- moved as it may disturb their mental health and may decrease their self confidence. During the last decade, hateful and sexist content towards women is being increasingly spread on social networks. The exposure to sexist speech has se- rious consequences to women’s life and limits their freedom of speech. Sexism is expressed in very different forms: it includes subtle stereo- types and attitudes that, although frequently un- noticed, are extremely harmful for both women and society. Sexist comments have a major im- pact on women being subjected to it. We as a team participated in the shared task Explain- able Detection of Online Sexism (EDOS) at Se- mEval 2023 and have proposed a model which identifies the sexist comments and its type from English social media posts using the data set shared for the task. Different tranformer mod- els like BERT, DistilBERT, and RoBERTa are used by the proposed model for implementing all the three tasks shared by EDOS. On using the BERT model, macro F1 score of 0.8073, 0.5876 and 0.3729 are achieved for Task A, Task B and Task C respectively. 1 Introduction There is a lot of information being spread in so- cial media. These can be categorised as necessary and unnecessary information. Any information that helps in the growth of a persona and society is con- sidered as necessary information. Any information that is in any way not useful to the society or harms the livelihood and mental health of an individual is considered as unnecessary information which can be removed for the benefit of the society. Hateful and sexist content towards women that are increasingly spread in the social media net- works has a serious impact on their life and lim- its their freedom of speech. An exhaustive analy- sis is required to understand how sexist behaviour and attitudes prevail in social media platforms (Rodríguez-Sánchez et al., 2020). Online violence against women can have a serious impact on them as it may affect their mental health or bring hesi- tation in voicing out their points in social media. Hence it is important to find whether the content that is present in social media is sexist and such content should be removed (Shimi et al., 2022). It is necessary to explain the reason behind the classification of text as sexist.Thus an …

Content truncated to 500 words