Guy Aglionby
I am a final year PhD student at the University of Cambridge Computer Lab, supervised by Prof Simone Teufel within the Natural Language and Information Processing group. I am also a member of Homerton College. Before this, I completed the MPhil in Advanced Computer Science here and did my BSc in Computer Science at the University of Bristol. My master's research, entitled Modelling polysemy in distributional semantic models, looked at how word sense can be explicitly modelled in DSMs and the benefits of doing so. You can download my CV here.
My PhD research is on developing interpretable models for common sense multi-hop reasoning. The premise of this research is that it's important to know which facts a model has used to come to a conclusion so developers can check whether the combination makes sense. I use a knowledge graph to structure these facts and evaluate models on a question answering task. Models typically consist of separate text and graph encoders; interpretability techniques can be applied to the latter to retrieve a graph-structured explanation. I have created an ontology for structuring common sense facts, and my current work is in evaluating whether we can evaluate model explanations against gold standards in this format. I am also investigating the collection of relevant data from an ontology for input into these models.
I am separately interested in how tools from NLP can be used to help researchers in other fields, for example by assisting with data coding or exploration of qualitative data. These tools also have various business use cases. Please reach out if this sounds interesting!
I used to jointly organise the weekly NLIP seminar series. All continue to be welcome to join, especially given the current online format!
Outside of the lab I enjoy Taekwondo and kickboxing, and was the president of the kickboxing society for 2019-20.
Publications
Faithful Knowledge Graph Explanations in Commonsense Question Answering (pdf)
Guy Aglionby, Simone Teufel
In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022)
Abu Dhabi, UAE, 2022
Identifying relevant common sense information in knowledge graphs (pdf)
Guy Aglionby, Simone Teufel
In Proceedings of the First Workshop on Commonsense Representation and Reasoning (CSRR 2022)
Dublin, Ireland, 2022
Using machine learning to create a repository of judgments concerning a new practice area: a case study in animal protection law (pdf)
Joe Watson, Guy Aglionby, Samuel March
Artificial Intelligence and Law (2022)
CAMsterdam at SemEval-2019 Task 6: Neural and graph-based feature extraction for the identification of offensive tweets (pdf)
Guy Aglionby, Christopher Davis, Pushkar Mishra, Andrew Caines, Helen Yannakoudakis, Marek Rei, Ekaterina Shutova and Paula Buttery
In Proceedings of the International Workshop on Semantic Evaluation 2019 (SemEval 2019)
Minneapolis, USA, 2019
Teaching
- Machine Learning and Real-world Data (Part IA/IB): LT21 (supervisor), LT20 (head demonstrator/ticker, supervisor), LT19 (demonstrator/ticker, supervisor)
- Natural Language Processing (Part II): MT19 (head demonstrator), MT18 (head demonstrator)
- Data Science: principles and practice (Part II): MT19 (ticker)
- Scientific Computing (Part IA): LT21 (supervisor), LT20 (ticker), LT19 (ticker)
- Overview of Natural Language Processing (MPhil): MT18 (demonstrator)
Key: MT = Michaelmas term, autumn. LT = Lent term, winter.
News
| Nov 2022 | I am giving a talk at ETH Zürich "Collecting Knowledge Graph Explanations for Commonsense QA via Counterfactual Annotation". |
| Oct 2022 | Our work on faithful explanations in KG-based QA systems was accepted to EMNLP. See you in Abu Dhabi! |
| May 2022 | I will be at ACL in Dublin, presenting work on identifying relevant common sense information in knowledge graphs. |
| May 2022 | Our work on identifying legal judgments that concern animal protection has been published in Artificial Intelligence and Law. |
| October 2020 | I'm a technical co-chair for the Conference for Truth and Trust Online 2020. |
| September 2020 | I'm interning as an Applied Scientist at Amazon Alexa this autumn. |
| July 2020 | Attended ACL 2020. |
| July 2020 | Recognised as a significant contributor to the ACL 2020 virtual conference website. |
| June 2020 | I made a tool to update BibTeX files with canonical information from the ACL Anthology. |
| June 2020 | Attended AKBC 2020. |
| June 2020 | Gave a talk at PyData Sofia: "Which laws should go? Helping social scientists analyse public opinion with natural language processing". |
| May 2020 | Attended ICLR 2020. |
Contact
Email: firstname.lastname@cl.cam.ac.uk
Phone: +44 (0)1223 763558
Department of Computer Science and Technology
University of Cambridge
15 JJ Thomson Avenue
Cambridge CB3 0FD
Last updated 2022/10/29