theses

I supervise BSc and MSc dissertations at Mediterranean College, usually on topics close to my own research: applied deep learning, medical AI, and AI for blockchain.

Looking for a topic for 2026–27? See the thesis proposals below.

MSc dissertations

Vasileios Bakalis — LLM-Based Classification of Move Smart-Contract Issues in the Sui Ecosystem: Dataset Consolidation and Baseline Model Comparison (2026) MSc in Big Data Analytics, Mediterranean College / University of Derby. Built the curated dataset of ~3,300 labelled Sui Move snippets and benchmarked LLM baselines; the dataset is archived on Zenodo and underpins the NiceMove VS Code extension. PDF

BSc dissertations

Vasileios Eirinaios Chanoutsos — Security Comparison of Smart Contracts in Sui vs. Other Blockchains (2026) BSc (Hons) Computer Science, Mediterranean College / University of Derby. Systematic security analysis comparing Move/Sui smart contracts against Solidity/EVM equivalents across multiple vulnerability classes. PDF

Dimitrios Zerkidis — AI-Based Code Style Enforcement with Code Llama (2026) BSc (Hons) Computer Science, Mediterranean College / University of Derby. Fine-tuned Code Llama for automated code style enforcement, evaluating adherence to style guides across Python codebases. PDF

Thesis proposals for 2026–27

Sui Move Thesis Proposals Topics offered this academic year, in collaboration with the Sui ecosystem. One research topic on evaluating the Move issue classifier beyond injected errors, and five data-collection topics that extend the Sui Move Issues dataset, one for each category: Perfect, Syntax Error, Semantic Error, Security Error and Style Error. Each topic has a theoretical and a practical component and is offered to one student only. The research topic and the Security Error topic have been allocated; the Perfect, Syntax Error, Semantic Error and Style Error topics are still available. PDF

Thesis Proposals in Artificial Intelligence for Cutaneous Lymphoma Research Nine BSc and MSc dissertation topics exploring AI for cutaneous lymphoma and treatment information, with input from medical researchers. Eight topics involve Python implementation and experiments in deep learning, vision-language models, retrieval and verification. One topic is a literature-based review of existing AI research and requires no programming. All topics use public datasets or accessible research literature, without requiring access to hospital records. The PDF explains each topic’s aims, practical work, expected outputs and required skills. Each topic is offered to one student only. Topics 1 and 5 have been allocated; topics 2, 3, 4, 6, 7, 8 and 9 are still available. PDF