A retrieval-augmented prompting network for hateful meme detection

The rise of user-generated content on social media is making memes a prevalent medium for expression. However, some memes convey offensive information toward individuals or groups on particular aspects. Detecting such harmful content is essential to mitigate potential conflicts and harm. This paper...

Full description

Saved in:
Bibliographic Details
Main Authors: Qiuhua Kuang, Yihao Lin, Junxi Liu, Xiazhi Lai, Runlu Zhong
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-07-01
Series:Frontiers in Physics
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fphy.2025.1614267/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849425693565255680
author Qiuhua Kuang
Yihao Lin
Junxi Liu
Xiazhi Lai
Runlu Zhong
author_facet Qiuhua Kuang
Yihao Lin
Junxi Liu
Xiazhi Lai
Runlu Zhong
author_sort Qiuhua Kuang
collection DOAJ
description The rise of user-generated content on social media is making memes a prevalent medium for expression. However, some memes convey offensive information toward individuals or groups on particular aspects. Detecting such harmful content is essential to mitigate potential conflicts and harm. This paper proposes a retrieval-augmented prompting network (RAPN) for hateful meme detection. The proposed model utilizes a retrieval-augmented selector to identify semantically relevant prompting examples from diverse sources, enhancing the selection to better match the inference instances. Based on the prompting framework, attention networks are employed to extract critical features from input instance and examples. By applying contrastive learning to label and feature spaces, the model is capable of learning more discriminative information for classification. Comprehensive evaluations on benchmark datasets demonstrate that our model outperforms the baseline methods. Thereby, the proposed model has strong evidence of high accuracy on the task of hateful meme classification.
format Article
id doaj-art-897ff8096d7f43569e12cc12b140b16a
institution Kabale University
issn 2296-424X
language English
publishDate 2025-07-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Physics
spelling doaj-art-897ff8096d7f43569e12cc12b140b16a2025-08-20T03:29:40ZengFrontiers Media S.A.Frontiers in Physics2296-424X2025-07-011310.3389/fphy.2025.16142671614267A retrieval-augmented prompting network for hateful meme detectionQiuhua Kuang0Yihao Lin1Junxi Liu2Xiazhi Lai3Runlu Zhong4School of Computer Science, Guangdong University of Education, Guangzhou, ChinaSchool of Electronic Science and Engineering, South China Normal University, Foshan, ChinaSchool of Electronic Science and Engineering, South China Normal University, Foshan, ChinaSchool of Computer Science, Guangdong University of Education, Guangzhou, ChinaSchool of Information Engineering, Guangzhou Panyu Polytechnic, Guangzhou, ChinaThe rise of user-generated content on social media is making memes a prevalent medium for expression. However, some memes convey offensive information toward individuals or groups on particular aspects. Detecting such harmful content is essential to mitigate potential conflicts and harm. This paper proposes a retrieval-augmented prompting network (RAPN) for hateful meme detection. The proposed model utilizes a retrieval-augmented selector to identify semantically relevant prompting examples from diverse sources, enhancing the selection to better match the inference instances. Based on the prompting framework, attention networks are employed to extract critical features from input instance and examples. By applying contrastive learning to label and feature spaces, the model is capable of learning more discriminative information for classification. Comprehensive evaluations on benchmark datasets demonstrate that our model outperforms the baseline methods. Thereby, the proposed model has strong evidence of high accuracy on the task of hateful meme classification.https://www.frontiersin.org/articles/10.3389/fphy.2025.1614267/fullhateful meme detectionpromptretrieval-augmented strategyattention mechanismcontrastive learning
spellingShingle Qiuhua Kuang
Yihao Lin
Junxi Liu
Xiazhi Lai
Runlu Zhong
A retrieval-augmented prompting network for hateful meme detection
Frontiers in Physics
hateful meme detection
prompt
retrieval-augmented strategy
attention mechanism
contrastive learning
title A retrieval-augmented prompting network for hateful meme detection
title_full A retrieval-augmented prompting network for hateful meme detection
title_fullStr A retrieval-augmented prompting network for hateful meme detection
title_full_unstemmed A retrieval-augmented prompting network for hateful meme detection
title_short A retrieval-augmented prompting network for hateful meme detection
title_sort retrieval augmented prompting network for hateful meme detection
topic hateful meme detection
prompt
retrieval-augmented strategy
attention mechanism
contrastive learning
url https://www.frontiersin.org/articles/10.3389/fphy.2025.1614267/full
work_keys_str_mv AT qiuhuakuang aretrievalaugmentedpromptingnetworkforhatefulmemedetection
AT yihaolin aretrievalaugmentedpromptingnetworkforhatefulmemedetection
AT junxiliu aretrievalaugmentedpromptingnetworkforhatefulmemedetection
AT xiazhilai aretrievalaugmentedpromptingnetworkforhatefulmemedetection
AT runluzhong aretrievalaugmentedpromptingnetworkforhatefulmemedetection
AT qiuhuakuang retrievalaugmentedpromptingnetworkforhatefulmemedetection
AT yihaolin retrievalaugmentedpromptingnetworkforhatefulmemedetection
AT junxiliu retrievalaugmentedpromptingnetworkforhatefulmemedetection
AT xiazhilai retrievalaugmentedpromptingnetworkforhatefulmemedetection
AT runluzhong retrievalaugmentedpromptingnetworkforhatefulmemedetection