Multitask Learning for Emotionally Analyzing Sexual Abuse Disclosures

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 4881–4892 June 6–11, 2021. ©2021 Association for Computational Linguistics 4881 Multitask Learning for Emotionally Analyzing Sexual Abuse Disclosures Ramit Sawhney†, Puneet Mathur‡, Taru Jain†, Akash Kumar Gautam†, Rajiv Ratn Shah† † Department of Computer Engineering, IIIT-Delhi {ramits, akash15011, rajivratn}@iiitd.ac.in ‡ University of Maryland, College Park puneetm@cs.umd.edu Abstract The #MeToo movement on social media plat- forms initiated discussions over several facets of sexual harassment in our society. Prior work by the NLP community for automated identification of the narratives related to sex- ual abuse disclosures barely explored this so- cial phenomenon as an independent task. How- ever, emotional attributes associated with tex- tual conversations related to the #MeToo so- cial movement are complexly intertwined with such narratives. We formulate the task of identifying narratives related to the sexual abuse disclosures in online posts as a joint modeling task that leverages their emotional attributes through multitask learning. Our results demonstrate that positive knowledge transfer via context-specific shared represen- tations of a flexible cross-stitched parame- ter sharing model helps establish the inher- ent benefit of jointly modeling tasks related to sexual abuse disclosures with emotion clas- sification from the text in homogeneous and heterogeneous settings. We show how for more domain-specific tasks related to sexual abuse disclosures such as sarcasm identifica- tion and dialogue act (refutation, justification, allegation) classification, homogeneous multi- task learning is helpful, whereas for more gen- eral tasks such as stance and hate speech de- tection, heterogeneous multitask learning with emotion classification works better.1 1 Introduction The #MeToo movement2 was started as an initia- tive to empower women against long-standing is- sues related to sexual abuse at workplaces, public spaces, and private organizations (McKenna and Chughtai, 2020). The usage of a dedicated hash- tag #MeToo on media platforms signified a social support system for women from different sections 1Code & Implementation: https://github.com/ midas-research/metoo-mtl-naacl 2https://metoomvmt.org/ Figure 1: Examples showing the relationship between tweets annotated for sexual harassment disclosure (top) and emotion recognition (bottom). Colors highlight to- ken level attention assigned by BERTweet. of society. The movement initiated discussions on many socially stigmatized issues that were missing from the virtual space (Clark-Parsons, 2019). Such conversations invited various reactions on the web, involving support to the cause of the movement and even outright bullying. While many users took part in the vilification of the survivors, the move- ment also saw opposition by factions of the society that felt threatened by the impact of social media in raising awareness about the scale of everyday sexual harassment faced by women in workplaces and institutions (Tambe, 2018). In many instances, the public disclosures of survivor-narrated inci- dents involved widespread use of hate-language and online trolling, both against the victims and alleged oppressors (Franks, 2019). The #MeToo movement also led to people coming out with alle- gations, refutations, and justifications about trau- matic experiences as they transitioned to active participants in the mainstream conversation (Gau- tam et …

Content truncated to 500 words