Home // HUSO 2019, The Fifth International Conference on Human and Social Analytics // View article

Measuring the Impact of Sentiment for Hate Speech Detection on Twitter

Authors:
Nina Bauwelinck
Els Lefever

Keywords: Hatespeech Detection; Sentiment Analysis; Twitter.

Abstract:
While social media platforms, such as Twitter offer users the opportunity to express their opinions and insights freely, there is a significant risk of users silencing each other based on prejudice by means of hateful Tweets. Since Twitter’s public nature makes these messages more widely disseminated, it is important to aid in the detection of such messages, which may cause harm to targeted (groups of) users. Following current state of the art, we assume the usefulness of sentiment features for the detection of hate speech messages, which tend to exhibit a higher degree of negative polarity. Therefore, we investigate the impact of these sentiment features as well as Twitter-specific and hate speech features on the performance of a supervised classification method with Support Vector Machines (SVMs). The Twitter-specific features offer the best performance increase over our strong token n-gram baseline.

Pages: 17 to 22

Copyright: Copyright (c) IARIA, 2019

Publication date: June 30, 2019

Published in: conference

ISSN: 2519-8351

ISBN: 978-1-61208-725-2

Location: Rome, Italy

Dates: from June 30, 2019 to July 4, 2019