SafeCity: Understanding Diverse Forms of Sexual Harassment Personal Stories
Sweta Karlekar, Mohit Bansal

TL;DR
This paper introduces models for automatically categorizing diverse sexual harassment stories from SafeCity, achieving high accuracy and providing interpretability to aid safety measures and incident reporting.
Contribution
It presents CNN-RNN models for harassment classification and interpretable analysis techniques to understand neural predictions in this context.
Findings
Single-label CNN-RNN achieves 86.5% accuracy
Multi-label model attains 82.5% Hamming score
Interpretability methods reveal features for safety applications
Abstract
With the recent rise of #MeToo, an increasing number of personal stories about sexual harassment and sexual abuse have been shared online. In order to push forward the fight against such harassment and abuse, we present the task of automatically categorizing and analyzing various forms of sexual harassment, based on stories shared on the online forum SafeCity. For the labels of groping, ogling, and commenting, our single-label CNN-RNN model achieves an accuracy of 86.5%, and our multi-label model achieves a Hamming score of 82.5%. Furthermore, we present analysis using LIME, first-derivative saliency heatmaps, activation clustering, and embedding visualization to interpret neural model predictions and demonstrate how this extracts features that can help automatically fill out incident reports, identify unsafe areas, avoid unsafe practices, and 'pin the creeps'.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsHate Speech and Cyberbullying Detection · Cancer-related gene regulation · Sexual Assault and Victimization Studies
MethodsLocal Interpretable Model-Agnostic Explanations
