This paper is a summary of the work done in my PhD thesis. Where I investigate the impact of bias in NLP models on the task of hate speech detection from three perspectives: explainability, offensive stereotyping bias, and fairness. Then, I discuss the main takeaways from my thesis and how they can benefit the broader NLP community. Finally, I discuss important future research directions. The findings of my thesis suggest that the bias in NLP models impacts the task of hate speech detection from all three perspectives. And that unless we start incorporating social sciences in studying bias in NLP models, we will not effectively overcome the current limitations of measuring and mitigating bias in NLP models.

In recent years, gray social media platforms, those with a loose moderation policy on cyberbullying, have been attracting more users. Recently, data collected from these types of platforms have been used to pre-train word embeddings (social-media-based), yet these word embeddings have not been investigated for social NLP related tasks. In this paper, we carried out a comparative study between social-media-based and non-social-media-based word embeddings on two social NLP tasks: Detecting cyberbullying and Measuring social bias. Our results show that using social-media-based word embeddings as input features, rather than non-social-media-based embeddings, leads to better cyberbullying detection performance. We also show that some word embeddings are more useful than others for categorizing offensive words. However, we do not find strong evidence that certain word embeddings will necessarily work best when identifying certain categories of cyberbullying within our datasets. Finally, We show even though most of the state-of-the-art bias metrics ranked social-media-based word embeddings as the most socially biased, these results remain inconclusive and further research is required.

It has become crucial to develop tools for automated hate speech and abuse detection. These tools would help to stop the bullies and the haters and provide a safer environment for individuals especially from marginalized groups to freely express themselves. However, recent research shows that machine learning models are biased and they might make the right decisions for the wrong reasons. In this thesis, I set out to understand the performance of hate speech and abuse detection models and the different biases that could influence them. I show that hate speech and abuse detection models are not only subject to social bias but also to other types of bias that have not been explored before. Finally, I investigate the causal effect of the social and intersectional bias on the performance and unfairness of hate speech detection models.

Systematic Offensive stereotyping (SOS) in word embeddings could lead to associating marginalised groups with hate speech and profanity, which might lead to blocking and silencing those groups, especially on social media platforms. In this [id=stk]work, we introduce a quantitative measure of the SOS bias, [id=stk]validate it in the most commonly used word embeddings, and investigate if it explains the performance of different word embeddings on the task of hate speech detection. Results show that SOS bias exists in almost all examined word embeddings and that [id=stk]the proposed SOS bias metric correlates positively with the statistics of published surveys on online extremism. We also show that the [id=stk]proposed metric reveals distinct information [id=stk]compared to established social bias metrics. However, we do not find evidence that SOS bias explains the performance of hate speech detection models based on the different word embeddings.