Fifteen-Plus Years, One Throughline

A Question That Kept Getting Bigger.

It started with a simple, almost stubborn idea: can a machine know how confident it should be? That question about uncertainty never left — it just found new places to live: in word embeddings, in hypergraphs running inside eBay's production stack, in retinal scans, and now in whether a language model understands the culture of the person talking to it.

0 Publications in Top Venues
0 Citations, Semantic Scholar
0 h-index
0 Institutions & Companies
The Throughline

How One Question Evolved.

Not a career timeline — that's a different story. This is the arc of the ideas themselves: the same instinct about uncertainty, followed through five very different disguises.

2008–2011 · Hong Kong
The Question

An unsupervised idea about ranking documents by difficulty — inspired, oddly, by Super Mario level design — becomes a PhD in latent probabilistic topic discovery, and a Best Paper award at the Beijing–Hong Kong Doctoral Forum.

2015–2019 · Cardiff & Kent
Making Meaning Measurable

Word embeddings stop being black boxes. Conceptual spaces, von Mises–Fisher priors, embeddings framed as principled Bayesian estimation — published at ACL, AAAI and IJCAI.

2019–2022 · Essex & Industry
Theory Meets Production

Multi-modal hypergraphs ship inside eBay's recommendation stack; a transformer model goes live inside BT's customer systems. The lab work starts paying rent in the real world.

2023–2025 · Southampton
Can We Trust It?

Bayesian uncertainty modelling, with Thomson Reuters Labs, asks whether AI should be willing to say "I don't know." The same instinct turns retinal scans into a biomarker for whole-body health — featured on Forbes.

2025– · Global
Whose Values?

CALM teaches language models cultural self-awareness. Policy work with the Royal Society, and a Ministry of Defence project with the Alan Turing Institute, ask what responsible AI actually requires.

Four Pillars

The Same Idea, Four Disguises.

Read top to bottom and you're reading the arc above, slowed down.

Pillar I

Probabilistic Foundations

Every model I've built since 2009 rests on the same conviction: intelligence isn't about being right, it's about knowing how unsure you are. My PhD work on latent topic discovery grew into a decade of Bayesian non-parametric models and approximate inference — the mathematical bedrock that lets AI reason honestly about what it doesn't know, in domains where a falsely confident answer is worse than none at all.

1,200+ citations h-index 18 Foundational since 2009
AAAI 2019

Word Embedding as Maximum A Posteriori Estimation

ACL 2019

Word and Document Embedding with vMF-Mixture Priors on Context Word Vectors

ECIR 2015

Nonparametric Topic Modeling using Chinese Restaurant Franchise with Buddy Customers

WSDM 2025 · with Thomson Reuters Labs

BAKER: Bayesian Kernel Uncertainty in Domain-Specific Document Modelling

Pillar II

Multi-modal Learning at Scale

Real data doesn't arrive as clean text — it's images, clickstreams and structured metadata, all at once. I build hypergraph models that capture relationships across these modalities simultaneously, work that moved from a whiteboard to production traffic at eBay and British Telecom, and picked up an eBay Top-3 Leaders' Choice Award along the way, selected out of 2,000+ global submissions.

Deployed at eBay & BT CIKM Best Paper Nominee eBay Top-3 Leaders' Choice
CIKM 2021 · Best Paper Nominee

Click-Through Rate Prediction with Multi-Modal Hypergraphs

IEEE Access 2021 · Deployed at BT

Unified Transformer Multi-task Learning for Intent Classification with Entity Recognition

AAAI 2022

Inferring Prototypes for Multi-Label Few-Shot Image Classification with Word Vector Guided Attention

WSDM 2025

GAMED: Knowledge Adaptive Multi-Experts Decoupling for Multimodal Fake News Detection

Pillar III

Digital Health & Oculomics

The eye turns out to be a remarkably honest witness to what's happening in the rest of the body. My work in Oculomics fuses computer vision with medical imaging to surface early biomarkers for cardiovascular and systemic disease — research that Forbes described as turning eye exams into a gateway to whole-body wellness, and that leans on Pillar I's uncertainty modelling to make medical predictions AI is willing to stand behind.

Featured on Forbes Most-cited paper: 175+ citations With Thomson Reuters Labs
Featured on Forbes, Nov 2025

The Eye as a Window to Systemic Health: A Survey of Retinal Imaging from Classical Techniques to Oculomics

World Wide Web Journal 2021 · Most-cited

Explainable Depression Detection with Multi-Modalities Using a Hybrid Deep Learning Model on Social Media

WISE 2024 · with Thomson Reuters Labs

Would You Trust an AI Doctor? Building Reliable Medical Predictions with Kernel Dropout Uncertainty

MICAD 2025

Retinal–Lipidomics Associations as Candidate Biomarkers for Cardiovascular Health

Pillar IV

Responsible AI & Governance

As LLMs get folded into everyday decisions, "it works on the benchmark" stops being good enough. My recent work makes models culturally self-aware and interrogates the bias baked into their architecture and training data — the same questions that led the Royal Society to pair me with a UK Home Office policymaker, and that now underpin a Ministry of Defence project on sensemaking under uncertainty with the Alan Turing Institute.

NeurIPS 2025 Royal Society Pairing Scheme MoD / Alan Turing Institute
NeurIPS 2025

CALM: Culturally Self-Aware Language Models

SIGIR 2025

Bias in Language Models: Interplay of Architecture and Data?

COLING 2025 Workshop

Evaluating Large Language Models on Health-Related Claims Across Arabic Dialects

COLING 2025

Uncertainty Modelling in Under-Represented Languages with Bayesian Deep Gaussian Processes

How It All Connects

One Foundation, Three Frontiers.

Hover a node. Nothing here is a coincidence — the uncertainty methods from Pillar I are load-bearing for everything downstream.

Probabilistic Foundations Multi-modal eBay · BT Digital Health Oculomics Responsible AI Governance

Hover or tap a node to see how it connects to the rest of the work.

Fifteen Years On, Still Being Built On

The Work Doesn't Stop at Publication.

A handful of real papers that cite this work — sourced from Semantic Scholar — chosen to show how far back some of it reaches.

A Comparative Study of Neural Sinkhorn Topic Models Based on Different Word Embeddings (2025)

Builds on Jointly Learning Word Embeddings and Latent Topics (SIGIR 2017), comparing it against newer optimal-transport-based topic models.

8-year gap

Multi-graph Collaborative and Global Hypergraph Sampling for Multimodal Recommendation (2026)

Directly extends the eBay-deployed Click-Through Rate Prediction with Multi-Modal Hypergraphs (CIKM 2021) with a new sampling strategy.

5-year gap

Dep-LLM: Training-Free Depression Diagnosis via Evidence-Guided Structured Multi-factor Reasoning (2026)

Cites the most-cited paper in this whole archive — Explainable Depression Detection with Multi-Modalities (2021, 175+ citations) — while exploring LLM-based diagnostic reasoning.

5-year gap

DeepQFM: A Deep Learning Based Query Facets Mining Method (2023)

Builds on Web Query Reformulation via Joint Modeling of Latent Topic Dependency (TOIS 2015) to mine query facets with deep learning.

8-year gap
Browse All 97 Publications → See the Full Career Story Google Scholar