Multi-Modal Discussion Transformer: Integrating Text, Images and Graph Transformers to Detect Hate Speech on Social Media
Abstract
We present the Multi-Modal Discussion Transformer (mDT), a novel method for detecting hate speech on online social networks such as Reddit discussions. In contrast to traditional comment-only methods, our approach to labelling a comment as hate speech involves a holistic analysis of text and images grounded in the discussion context. This is done by leveraging graph transformers to capture the contextual relationships in the discussion surrounding a comment and grounding the interwoven fusion layers that combine text and image embeddings instead of processing modalities separately. To evaluate our work, we present a new dataset, HatefulDiscussions, comprising complete multi-modal discussions from multiple online communities on Reddit. We compare the performance of our model to baselines that only process individual comments and conduct extensive ablation studies.
Cite
Text
Hebert et al. "Multi-Modal Discussion Transformer: Integrating Text, Images and Graph Transformers to Detect Hate Speech on Social Media." AAAI Conference on Artificial Intelligence, 2024. doi:10.1609/AAAI.V38I20.30213Markdown
[Hebert et al. "Multi-Modal Discussion Transformer: Integrating Text, Images and Graph Transformers to Detect Hate Speech on Social Media." AAAI Conference on Artificial Intelligence, 2024.](https://mlanthology.org/aaai/2024/hebert2024aaai-multi/) doi:10.1609/AAAI.V38I20.30213BibTeX
@inproceedings{hebert2024aaai-multi,
title = {{Multi-Modal Discussion Transformer: Integrating Text, Images and Graph Transformers to Detect Hate Speech on Social Media}},
author = {Hebert, Liam and Sahu, Gaurav and Guo, Yuxuan and Sreenivas, Nanda Kishore and Golab, Lukasz and Cohen, Robin},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2024},
pages = {22096-22104},
doi = {10.1609/AAAI.V38I20.30213},
url = {https://mlanthology.org/aaai/2024/hebert2024aaai-multi/}
}