Hate Speech in Pixels: Detection of Offensive Memes towards Automatic Moderation

Auteur-es

Benet Oriol Sabat, Cristian Canton Ferrer, Xavier Giro-i-Nieto

Nombre Auteurs

3

Titre

Hate Speech in Pixels: Detection of Offensive Memes towards Automatic Moderation

Année de publication

2019

Référence (APA)

Sabat, B. O., Ferrer, C. C., & Giro-i-Nieto, X. (2019). Hate Speech in Pixels : Detection of Offensive Memes towards Automatic Moderation.

résumé

This work addresses the challenge of hate speech detection in Internet memes, and attempts using visual information to automatically detect hate speech, unlike any previous work of our knowledge. Memes are pixel-based multimedia documents that contain photos or illustrations together with phrases which, when combined, usually adopt a funny meaning. However, hate memes are also used to spread hate through social networks, so their automatic detection would help reduce their harmful societal impact. Our results indicate that the model can learn to detect some of the memes, but that the task is far from being solved with this simple architecture. While previous work focuses on linguistic hate speech, our experiments indicate how the visual modality can be much more informative for hate speech detection than the linguistic one in memes. In our experiments, we built a dataset of 5,020 memes to train and evaluate a multi-layer perceptron over the visual and language representations, whether independently or fused.

URL

https://research.facebook.com/file/678017116419206/Hate-Speech-in-Pixels-Detection-of-Offensive-Memes-towards-Automatic-Moderation.pdf

doi

https://doi.org/10.48550/arXiv.1910.02334

Accessibilité de l'article

Libre

Champ

Computer Vision

Type contenu (théorique Applicative méthodologique)

Applicatif

Méthode

The proposed method involves using a multi-layer perceptron model to detect hate speech in internet memes by analyzing both visual and linguistic information.

Cas d'usage

N/A

Objectifs de l'article

The objectives of the article are to address the harmful impact of hate memes and propose a solution for automatically detecting them.

Question(s) de recherche/Hypothèses/conclusion

The research question is how to effectively detect hate speech in internet memes using both visual and linguistic information.

The hypothesis is that a multi-layer perceptron model can effectively detect hate speech in internet memes by analyzing both visual and linguistic information.

The conclusions are that the proposed multi-layer perceptron model is effective in detecting hate speech in internet memes, and that future research should consider additional contextual information to improve detection accuracy.

Cadre théorique/Auteur.es

The theoretical framework of the article includes previous research on hate speech detection and multimodal analysis, with main authors cited including Z. Waseem, T. Davidson, and J. Pennington.

Concepts clés

Hate speech, Memes, Visual information, Linguistic information, Multi-layer perceptron model.

Données collectées (type source)

Internet memes from various social media platforms, and used OCR technology to extract text from the images.

Définition des émotions

Non

Ampleur expérimentation (volume de comptes)

5000 memes

Technologies associées

CR technology for extracting text from images, and a multi-layer perceptron model for analyzing both visual and linguistic information.

Mention de l'éthique

Non