Jonathan Sumner

Research Overview

My PhD, supervised by Professor Vaclav Brezina (ESRC Centre for Corpus Approaches to Social Science), combines learner corpus research with AI. I am building the Korean Multimodal English Corpus (K-MEC), a longitudinal corpus of spoken and written English from Korean learners, to investigate how grammatical accuracy differs between speech and writing.

The corpus also underpins an AI teaching system: using supervised fine-tuning and retrieval-augmented generation (RAG), I aim to show that a model grounded in learner data can give more transparent, pedagogically informed feedback than general-purpose chatbots.

I have developed an ASR-assisted dashboard for verbatim transcription to the Lancaster standard.