Investigating the Effectiveness of Culturally Related Computational Humour and Empathetic Functions in English as an Additional Language (EAL) Acquisition in a Digital Environment

Approximately 750 million people speak English as a second or additional language (EAL) learning, and it is anticipated that the number of people studying English in different parts of the globe will continue to rise. According to research conducted by the British Council, there are presently 1.75 billion individuals worldwide who speak English, which equates to one in four people globally, and it is predicted that by the year 2020, there will be 2 billion individuals utilising the language. Advancements in digital technologies have significantly transformed the landscape of EAL education. With AI integration, chatbots designed for EAL have become crucial tools. These chatbots facilitate human-chatbot interactions that enhance the learning experience by creating more engaging and efficient environments, providing personalised teaching and immediate feedback, as well as analytics and diagnostic tools. Collectively, these features revolutionise EAL learning by tailoring education to meet individual needs. The use of AI in EAL education, however, faces several significant challenges. First, the effectiveness of AI tools heavily relies on the learner's self-discipline and motivation, which can vary widely among individuals and may lead to inconsistent engagement and outcomes. Second, the repetitiveness of many AI-driven educational programs can result in boredom and disengagement, especially if the learning activities lack sufficient variety or fail to match the learner's interests and learning style. Lastly, AI systems may struggle to effectively address the unique obstacles learners encounter, as they lack the nuanced understanding and adaptability of human teachers, potentially leading to frustration and hindered progress in learners who face significant learning challenges. Building on these challenges, existing chatbots designed for EAL acquisition have specific shortcomings in intuitively recognising and responding to learners' emotional needs and the difficulties prompted by the learning process. These chatbots often fail to intuitively identify these needs and difficulties incited by learning, offer empathetic support, or reduce learning boredom through appropriate humour that considers learners' cultural backgrounds. While a few studies have acknowledged the significance of humour, empathy and cultural elements in EAL education, these aspects have been explored in isolation – either focusing on humour without consideration of culture or on empathy without accounting for these backgrounds. An integrated model that synthesises these elements to comprehensively cater to EAL learners' unique circumstances has yet to be broadly implemented. This study develops the Multi Adversarial Network Embedded with Culture, Humour, and Empathy (MACHE-Bot), utilising the Wasserstein Generative Adversarial Network (WGAN) model. MACHE-Bot is designed to intuitively identify EAL learners’ difficulties, offering culturally relevant empathetic encouragement and incorporating culturally appropriate humour to stimulate curiosity, enhance motivation, and improve learning retention. The key findings of the MACHE-Bot are (a) enhanced engagement and motivation, (b) improved retention, (c) enhanced cultural understanding, (d) relevance of culturally infused humour, and (e) relevance of culturally infused empathy. The practical outcomes of this research offer valuable insights for EAL learners, educators, and technologists into designing emotionally and culturally responsive chatbot technologies. These innovations promise more engaging and effective learning experiences for EAL learners. By diminishing language barrier-related anxiety and boosting motivation, these applications have the potential to significantly transform EAL education, making it more accessible and successful for learners across the globe.