Abstract
The rise of social media platforms has significantly changed the way our world communicates, and part of those changes includes a rise in inappropriate behaviors, such as the use of aggressive and hateful language online. Detecting such content is crucial to filtering or blocking inappropriate content on the Web. However, due to the huge amount of data posted every day, automatic methods are essential for identifying this type of content. Seeking to address this issue, the Natural Language Processing community is increasingly involved in testing a wide range of techniques for hate speech detection. While achieving promising results, these techniques consider hate speech detection as the sole optimization objective, without involving other related tasks such as polarity and emotion classification that are strongly linked to offensive behavior. In this paper, we propose the first Multi-task approach that leverages the shared affective knowledge to detect hate speech in Spanish tweets, using a well-known Transformer-based model. Our results show that the combination of both polarity and emotional knowledge helps to detect hate speech more accurately across datasets.
Highlights
In recent decades, people have been getting hooked on social media platforms as a way of relating and connecting to other people
Since we are dealing with Twitter data, we perform a Twitterspecific data cleaning before including the tweets in the models
The spread of Hate Speech (HS) has increased in recent years becoming a major challenge for online platforms and national governments, which need to rely on automated systems to identify and remove this type of content
Summary
People have been getting hooked on social media platforms as a way of relating and connecting to other people. According to a psychological study [3], negative sentiment messages are often indicators of emotions, such as anger, disgust, fear or sadness, and positive texts are related to joy. Negative sentiments and emotions like anger, disgust, fear and sadness are presented in HS messages as a number of studies have already revealed in recent years [5], [6]. Given the influence of emotions in HS messages, in this study we investigated new ways of unearthing HS by modeling polarity and emotion analysis along with the HS detection task. We perform an error analysis in order to gain deeper insight about the proposed MTL model performance This analysis allows us to identify some peculiarities of Spanish-speaking users while expressing HS.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
