Jason Dillon

Biography

I am Jason Dillon, currently pursuing a PhD in Law and Computer Science at Keele University. I hold an LLB (Hons) from Staffordshire University and an LLM in AI and New Technologies from Keele University. My research focuses on the intersection of artificial intelligence and legal education, aiming to explore how AI can transform legal teaching and practice to better prepare future generations for a tech-driven world. My work experience includes roles as a Commercial Legal Advisor at Keele University Commercial Law Clinic and as a Governor at ESPRIT Multi Academy Trust, where I developed a deep understanding of regulatory and ethical issues related to emerging technologies. Additionally, I have served as a Research Assistant at Keele University, contributing to projects investigating the use of AI in legal education. Beyond my legal expertise, I have a solid foundation in programming, including Python, SQL, PHP, and JavaScript, which allows me to bridge the gap between traditional legal education and modern technological demands. My goal is to develop a tech-enhanced legal educational framework that integrates AI, ensuring the next generation of lawyers is well-equipped for the complexities of AI and digital transformation. I am passionate about innovation at the intersection of law and technology, and I actively engage with academic peers and industry experts to advance research in both fields. Ultimately, I aim to contribute to a more integrated, AI-driven legal curriculum that fosters both technological proficiency and traditional legal skills.

Research and scholarship

Working Title: "Leveraging Transformer-Based NLP Models for Enhanced Automated Grading Systems in Legal Education"

Objectives:

1. Develop an Automated Grading System: Build a robust NLP-based grading tool using transformer models like BERT or GPT, specifically tailored for evaluating complex legal essays and responses.
2. Ensure Accuracy and Fairness: Optimize the model to deliver grading results that align with human evaluators, ensuring consistency, reliability, and fairness across diverse student populations.
3. Enhance Explainability and Feedback: Integrate an explainability module to provide students and educators with a clear understanding of grading decisions, along with constructive feedback for learning improvement.
4. Mitigate Bias: Identify potential biases and implement mitigation strategies within the model to ensure equitable assessments for all students.
5. Evaluate Educational Impact: Assess the tool's impact on grading efficiency, instructor workload, and student learning outcomes.

Significance:

This research addresses the increasing demand for scalable, fair, and efficient grading systems in higher education, especially in fields that require complex analysis, such as law. By utilising advanced NLP models, this project aims to:

• Promote Educational Equity: Deliver consistent grading that is free from human fatigue or subconscious biases.
• Enhance Learning: Offer students timely and detailed feedback to enrich their educational experience.
• Support Instructors: Alleviate the grading workload for instructors, enabling them to dedicate more time to meaningful aspects of teaching and curriculum development.
• Advance Assessment Methods: Contribute to the evolving landscape of AI-driven education, with a particular focus on addressing the unique challenges of legal education.

Overall, this project seeks to push the boundaries of AI in educational assessment while establishing a framework for creating ethical and effective automated grading tools.