NLP and graph representation learning
Current work on language models, graph-based reasoning, and reproducible training and evaluation workflows in Python and PyTorch.
John Stewart Fabila-Carrasco
Postdoctoral Researcher in NLP and Graph Representation Learning
PhD in Mathematical Engineering
I am a postdoctoral researcher at Cardiff University working on graph-based machine learning, spectral graph theory, and graph signal processing. My research connects mathematical methods with NLP, biomedical signals, and public-sector data analysis.
15Peer-reviewed papers
207Google Scholar citations
23Research talks
250+Teaching contact hours
Current and recent
- 2025-2027Postdoctoral research at Cardiff University
NLP and graph representation learning, with Python and PyTorch workflows.
- 2025Signed Laplacians for Constrained Graph Clustering
International Conference on Machine Learning (ICML).
- 2024Graph-based analysis of task-related fMRI signals
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).
Academic path
- 2010-2020Mathematics and mathematical engineering
Training at UAEM, UNAM, and Universidad Carlos III de Madrid.
- 2020-2025Postdoctoral research at the University of Edinburgh
Graph signal processing, biomedical data, and scalable graph clustering.
- 2025-2027Current postdoctoral appointment at Cardiff University
NLP and graph representation learning.
ResearchAreas of work
Spectral graph theory and clustering
Magnetic Laplacians, spectral bracketing, isospectral graphs, and graph clustering methods for structured data.
Graph signal processing and biomedical data
Entropy and complexity measures for EEG, fMRI, DTI networks, and other noisy multivariate signals.
Public-sector data science
Statistical modelling, GIS equity analysis, reproducible R workflows, and Shiny dashboards for policy-facing work.
Academic appointmentsResearch experience
Sep 2025 - Apr 2027
Postdoctoral Researcher in NLP and Graph Representation Learning
Cardiff University · Cardiff, Wales, UK
Developing NLP methods that combine language modelling with graph reasoning, supported by GPU-accelerated training and evaluation workflows in Python and PyTorch.
Feb 2024 - Apr 2025
Postdoctoral Researcher in Scalable Spectral Clustering
University of Edinburgh, School of Informatics · Edinburgh, Scotland, UK
Designed clustering algorithms for large graph-structured datasets, with a focus on computational efficiency and scalable experimentation for high-dimensional network analysis.
Nov 2020 - Jan 2024
Postdoctoral Researcher in Graph Signal Processing for Biomedical Data
University of Edinburgh, School of Engineering · Edinburgh, Scotland, UK
Developed nonlinear graph-signal methods for EEG and neuroimaging time series, leading the computational analysis work for a Leverhulme-funded project on difficult healthcare data.
Feb 2014 - Sep 2015
Industry and Government Data Science Researcher
National Institute of Educational Evaluation (INEE) · Mexico City, Mexico
Led end-to-end statistical modelling for nationwide assessment and census data, building reproducible R pipelines and Shiny dashboards for policy and resource-allocation decisions.
Selected papersPublications
- 01
- 02
- 03
- 04
- 05
TeachingTeaching and supervision
Teaching experience across mathematics, statistics, and engineering in the UK and Spain, including 250+ contact hours and three MSc co-supervisions.
Contact
Research correspondence and academic collaboration
For research questions, talks, or collaboration, email is the best way to get in touch.