The Use of a Large Language Model for Cyberbullying Detection

Abstract

The dominance of social media has added to the channels of bullying for perpetrators. Unfortunately, cyberbullying (CB) is the most prevalent phenomenon in today’s cyber world, and is a severe threat to the mental and physical health of citizens. This opens the need to develop a robust system to prevent bullying content from online forums, blogs, and social media platforms to manage the impact in our society. Several machine learning (ML) algorithms have been proposed for this purpose. However, their performances are not consistent due to high class imbalance and generalisation issues. In recent years, large language models (LLMs) like BERT and RoBERTa have achieved state-of-the-art (SOTA) results in several natural language processing (NLP) tasks. Unfortunately, the LLMs have not been applied extensively for CB detection. In our paper, we explored the use of these models for cyberbullying (CB) detection. We have prepared a new dataset (D2) from existing studies (Formspring and Twitter). Our experimental results for dataset D1 and D2 showed that RoBERTa outperformed other models.

Natural Language Understanding for Navigation of Service Robots in Low-Resource Domains and Languages: Scenarios in Spanish and Nahuatl

“…Nowadays, it is a common practice to feed neural networks with word vector representations, using them as the first data-processing layer [43]. Furthermore, using pre-trained word embeddings has shown to be a powerful tool for different natural language processing tasks [44,45].…”

Higher education assessment practice in the era of generative AI tools

“…These tools were a result of innovative solutions developed using large language models (LLMs) such as the generative pre-trained transformers (GPT) series developed by OpenAI, generalised autoregressive pretraining for language understanding (XLNet) developed by Google, Salesforce's conditional transformer language models (CTRL), Google's Pathways language model (PaLM), and Meta's large language model meta-AI (LLaMa). The LLMs have been used for various natural language processing tasks such as questioning and answering (Pochiraju et al, 2023), sentiment analysis (Habbat et al, 2022), topic modelling , cyberbullying detection (Ogunleye & Dharmaraj, 2023), and fake news detection (Caramancion, 2023). As detailed in Table 1 below, the development of LLMs is mainly dominated by a few large organisations, including Google, Meta, and Microsoft/OpenAI.…”

A Systematic Review of Generative AI for Teaching and Learning Practice

“…By processing a huge amount of data, LLMs learn the statistical relationships, patterns, and structure within datasets, which enables them to predict or generate relevant and meaningful content in response to user requests. Thus, they are capable of performing various complex tasks [6][7][8][9].…”

“…For example, ChatGPT relies on the GPT series to perform its task. LLM s are trained on a large number of parameters (data), including text and images [5][6][7]. By processing a huge amount of data, LLMs learn the statistical relationships, patterns, and structure within datasets, which enables them to predict or generate relevant and meaningful content in response to user requests.…”

“…By processing a huge amount of data, LLMs learn the statistical relationships, patterns, and structure within datasets, which enables them to predict or generate relevant and meaningful content in response to user requests. Thus, they are capable of performing various complex tasks [6][7][8][9]. However, concerns including hallucinations [10], bias [11], ethical and privacy concerns [12,13], accidental plagiarism [14], and academic integrity [15][16][17][18][19][20] have been raised regarding Gen AI tools; such tools have been praised for their potential benefits in relation to H E .…”