Content Details | NUML Online Research Repository

Title UNRAVELING LINGUISTIC CHALLENGES: EXPLORING THE ROLE OF ARTIFICIAL INTELLIGENCE IN ONLINE HATE SPEECH AND HARASSMENT

Author(s) Shaista Noor

Abstract Title: Unraveling Linguistic Challenges: Exploring the Role of Artificial Intelligence in Online Hate Speech and Harassment The present research explores the role of Artificial Intelligence (AI) in the detection of online hate speech and the linguistic challenges encountered during the process. Grounded in Socio-Technical Systems Theory (STS) and Discourse Ethics Theory, the study investigates the linguistic challenges and ethical issues encountered by AI systems in identifying hate speech across diverse linguistic and cultural contexts. The research employed a mixed-method approach combining both quantitative and qualitative analyses. For the quantitative phase, the data was collected from online available datasets on websites such as Kaggle and Google Data Search. The analysis provided linguistic features and patterns of online hate speech on online platforms. It revealed that Twitter is the most widely used online platform for the spread of hate speech. Moreover, the analysis measured the frequency and percentage distribution of hate speech and confirmed that political hate speech is the most prevalent, followed by racism and religious hate speech. For the qualitative phase, interviews were conducted with 10 AI experts, working in different institutions. The interviews revealed several linguistic and ethical challenges faced by the AI models while detecting online hate speech. Some of these include the complexity of hate speech, lack of diversity of datasets on which the models are trained and the lack of contextual understanding. The present research contributes to the field of linguistics by advocating ethical AI systems and providing future recommendations for researchers and stakeholders. The findings underscore the significance of AI collaboration in ensuring transparency, and in tackling the evolving and complex nature of online hate speech. By analyzing the linguistic and ethical challenges, the research paves the way for more inclusive and effective AI systems, ultimately contributing to equitable and safer online environments.

Type Thesis/Dissertation

Faculty Languages

Department English

Language English

Publication Date 2025-10-28

Subject

Publisher

Contributor(s)

Format

Identifier

Source

Relation

Coverage

Rights