Current Research Activities

At present, I am active in the following research projects:

• The GEM Benchmark Project — Investigating the evaluation of large language models.
• The ReproHum Project — Exploring the reproducibility of human evaluations in Natural Language Generation.

In addition to the above projects, I also participate in reviewing and writing of research papers for conferences such as ACL, INLG, EMNLP, and COLING. For a list of authored or co-authored papers please see the publications section.

I was also the lead program chair for the 17th International Conference on Natural Language Generation in Tokyo, Japan. I helped to organise the conference by sending out the call for papers, recruiting reviewers, defining the conference program, communicating with authors, coordinating with local chairs, etc.

Past Research Work

Past research work at the University of Aberdeen involved the BabyTalk-Family (BT-Family) project. This used Natural Language Generation (NLG) technology to convert large sets of clinical neonatal patient data into automatic daily summaries for parents of babies in neonatal intensive care (NICU). This was done by using artificial intelligence and natural language technology to automatically analyse and summarise the information in a baby’s electronic patient record to produce reports focusing on matters of most interest to the parent. This work was based off a prototype that I developed during my PhD. BabyTalk-Family was successfully deployed and used by parents at the neonatal unit in Royal Infirmary of Edinburgh hospital during the research period.