Research

From graphs to knowledge models.

Explore the ideas behind the figures, the original papers and their connection to finance.

01 / Contextual learning

Learning from graph context.

KGPFN combines local neighbourhoods, relation patterns and example triples to predict missing links through in-context learning.

Local contextGlobal contextLink prediction

KGPFN architecture: local and global graph context feed a graph encoder, feature fusion and a prior-data fitted network. — Gao et al. · 2026 · Figure 2

02 / Knowledge refinement

Knowledge that can be revised.

DeepRefine uses answerability checks, error diagnosis and refinement actions to update an existing knowledge base as questions arrive.

DeepRefine Figure 3: a frozen policy processes successive queries and applies refinement actions to an evolving knowledge base. — Huang et al. · 2026 · Figure 3, v1

03 / Neural graph reasoning

Queries in graph and vector space.

A neural graph database encodes graph structure and queries into embedding space. The figure compares this path with symbolic database execution.

Figure 7 compares neural graph database retrieval in embedding space with symbolic query trees, joins and database indexes. — Ren et al. · 2023 · Figure 7

PRESENTATION CONTEXT

From Foundation Knowledge Graph Construction to Foundation Knowledge Models
Yangqiu Song · Department of CSE, HKUST / KnowComp
Original papers and authors are credited with each figure.

Platform architecture

Our technology stack.

05Applications

04AI Agents & Intelligence

03Financial Knowledge Layer

02Data & Model Infrastructure

01Security & Governance

Financial AI

Models and agents adapted to financial context.

Knowledge Intelligence

Graphs and retrieval connect facts to sources.

Data Intelligence

Market, document and portfolio data unified.

AI Governance

Access, audit and attribution integrated by design.