Dr Sha Liu · Presentations

Talks on AI-mediated language assessment — bookmark to keep an eye on new ones.
31 October 2026 · NIE ELL Seminar 2026, closing panel on GenAI in language learning and assessment · National Institute of Education, Singapore (recorded)

Prompted but misheard: How to provide personal and accurate speaking feedback in current spoken dialogue systems

Where ASR errors enter spoken dialogue systems, what they do to feedback that quotes the learner, and how far prompt design and model choice can contain them — with a study of 94 dialogues by Grade 8 learners in Vietnam.
20 August 2026 · LTRC Language Assessment Seminar Series · University of Melbourne (online)

Prompted but misheard: Construct realisation and the human-in-the-loop in AI-mediated L2 speaking

The ASR beneath speaking feedback — what the recogniser mishears, what a governed prompt fixes, and what only humans can catch.
21 May 2026 · Human-in-the-Loop Symposium · University of Cambridge

What are we assessing when AI plays multiple roles in the loop?

Human–AI collaboration in language assessment: constructs, prompting, and pipeline architectures.
2026 · LTRC · work in progress

Examining L2 learners' cognitive processing

Eye-tracking and learner engagement in AI-mediated assessment.
2026 · workshop

AI for a better workflow

With Dan — practical AI-assisted research workflows.
2026 · open call

CLARK field guides — a call for collaborators

Expert-curated, citable domain guides for AI-assisted language-assessment research.