I am currently working as a Research Scientist on DynAIRx (Artificial Intelligence for dynamic prescribing optimisation and care integration in multimorbidity) and MRIC (Mental Health for Research Innovation Center), both health data science projects hosted at the Civic Health Innovation Labs based in Liverpool. Previously, I worked as a Natural Language Processing Research Scientist at NEC Laboratories Europe (Heidelberg).

I am also a member of the the ECR supported by the AIM-RSF under the Alan Turing Institute.

Besides Machine learning, Data Science and Artificial Intelligence, I have in the past had experiences in software, networks and systems administration. I recently completed (2022) my PhD in Machine Learning and Natural Language Processing (NLP) specializing in Biomedical Information Extraction at the University of Liverpool working with Danushka Bollegala, Paula Williamson and Susanna Dodd.

Research Interests

My research interests span applications and topics pertaining to Machine learning and Natural language processing (NLP). I have particularly worked on problems such as information extraction, text generation and others with an objective of simplifying the way humans understand and communicate through written text. I’m keen on Representation learning, Few-shot & Zero-shot learning, Multi-task learning, Natural Language Generation, Non-autoregressive language generation, Entity recognition, Prompt based learning, Information retrieval, Multi-label text classification etc.

Recent works

▪ “Improving Pre-trained Language Model Sensitivity via Mask Specific losses: A case study on Biomedical NER” In 2024 NAACL, June, 2024, Micheal et al, pdf

▪ “Select and Augment: Enhanced Dense Retrieval Knowledge Graph Augmentation” Mar 2024, Proceedings of the AAAI Conference on Artificial Intelligence, Micheal et al, pdf.

▪ “Position-based Prompting for health outcome generation” In Biomedical Natural Language Process, 2023 ACL, May, 2022, Micheal et al, pdf

▪ “Detect and Classify–Joint Span Detection and Classification for Health Outcomes”, 2021 EMNLP Nov 2021 Micheal et al, pdf

▪ “Assessment of contextualised representations in detecting outcome phrases in clinical trials” In European Journal for Biomedical Informatics, Micheal et al pdf

▪ “Correcting crowdsourced annotations to improve detection of outcome-types in Evidence-Based Medicine.” In KDH@IJCAI, pp 1-5. 2019. Micheal et al pdf

Talks and Events

• Medical Research council (MRC) Clinical training fellowship teaching on AI and its applications in health care (April 8 2022)

• Monthly talks on Natural Language Processing applications, University of Liverpool, Jan 2019 – Present

Previous Responsibilities

Taken on duties in a Teaching Assistant role on Computer Science Modules in University of Liverpool

▪ Data Mining & Visualization (2019-2021)

▪ Machine Learning and Bioinspired Optimization (2022)

▪ Big Data Analysis (2019-2020)

▪ Data Structures and Algorithms (2019-2022)

▪ Scripting Languages (2019-2022)

▪ Introduction to Artificial Intelligence (2021-2022)

Honors and Awards

▪ PhD Studentship, Deep Learning (Biomedical Natural Language Processing), 2018 – 2021

▪ MSc Scholarship, Data Science and Analytics, Cardiff University, 2016 - 2017

▪ Best Oral Presentation at the MSc Student and Modelling Fellow Programme 2017 Showcase Event, Aneurin Bevan University Health Board (NHS Wales).

▪ Active Volunteer of the Mozilla project 2015 a part of the Mozilla Reps Program (2012-2016) under Mozilla Foundation.

▪ Recognition certificate for project exhibition during the Makerere University College of Information Sciences Open Day Exhibition.

Professional

Reviewer - Association for Computational Linguistics (ACL) Rolling Review (2021 - Present)

Others

Freelance Data Scientist

Contacts

micheal.abaho@neclab.eu

m.abaho@liverpool.ac.uk

michealabaho265@gmail.com

+44 7491574501

Full CV: available here