AOMD: An analogy-aware approach to offensive meme detection on social media

This paper focuses on an important problem of detecting offensive analogy meme online social media where the visual content and texts/captions together make to convey information. Existing detection solutions often ignore implicit relation between textual contents are insufficient identify memes. Two challenges exist in accurately memes: i) it is not trivial capture that implicitly conveyed by a meme; ii) also challenging effectively align complex across different data modalities meme. To address above challenges, we develop deep learning based Analogy-aware Offensive Meme Detection (AOMD) framework learn from multi-modal detect We evaluate AOMD two real-world datasets media. Evaluation results show achieves significant performance gains compared state-of-the-art baselines memes more accurately.