Toxic comment detection in online discussions
- Comment sections of online news platforms are an essential space to express opinions and discuss political topics. In contrast to other online posts, news discussions are related to particular news articles, comments refer to each other, and individual conversations emerge. However, the misuse by spammers, haters, and trolls makes costly content moderation necessary. Sentiment analysis can not only support moderation but also help to understand the dynamics of online discussions. A subtask of content moderation is the identification of toxic comments. To this end, we describe the concept of toxicity and characterize its subclasses. Further, we present various deep learning approaches, including datasets and architectures, tailored to sentiment analysis in online discussions. One way to make these approaches more comprehensible and trustworthy is fine-grained instead of binary comment classification. On the downside, more classes require more training data. Therefore, we propose to augment training data by using transfer learning. WeComment sections of online news platforms are an essential space to express opinions and discuss political topics. In contrast to other online posts, news discussions are related to particular news articles, comments refer to each other, and individual conversations emerge. However, the misuse by spammers, haters, and trolls makes costly content moderation necessary. Sentiment analysis can not only support moderation but also help to understand the dynamics of online discussions. A subtask of content moderation is the identification of toxic comments. To this end, we describe the concept of toxicity and characterize its subclasses. Further, we present various deep learning approaches, including datasets and architectures, tailored to sentiment analysis in online discussions. One way to make these approaches more comprehensible and trustworthy is fine-grained instead of binary comment classification. On the downside, more classes require more training data. Therefore, we propose to augment training data by using transfer learning. We discuss real-world applications, such as semi-automated comment moderation and troll detection. Finally, we outline future challenges and current limitations in light of most recent research publications.…
Author details: | Julian RischGND, Ralf KrestelORCiDGND |
---|---|
DOI: | https://doi.org/10.1007/978-981-15-1216-2_4 |
ISBN: | 978-981-15-1216-2 |
ISBN: | 978-981-15-1215-5 |
ISSN: | 2524-7565 |
ISSN: | 2524-7573 |
Title of parent work (English): | Deep learning-based approaches for sentiment analysis |
Publisher: | Springer |
Place of publishing: | Singapore |
Editor(s): | Basant Agarwal, Richi Nayak, Namita Mittal, Srikanta Patnaik |
Publication type: | Article |
Language: | English |
Date of first publication: | 2020/01/25 |
Publication year: | 2020 |
Release date: | 2022/11/10 |
Tag: | deep learning; hate speech detection; natural language processing; toxic comment classification; user-generated content |
Number of pages: | 25 |
First page: | 85 |
Last Page: | 109 |
Organizational units: | Digital Engineering Fakultät / Hasso-Plattner-Institut für Digital Engineering GmbH |
DDC classification: | 0 Informatik, Informationswissenschaft, allgemeine Werke / 00 Informatik, Wissen, Systeme |
Peer review: | Referiert |
License (English): | Creative Commons - Namensnennung-Weitergabe zu gleichen Bedingungen 3.0 Unported |