Data, Intelligence & Graphs – Télécom Paris

Home

The research interests of the DIG team span knowledge graphs, language models, foundational models, learning over tabular data, graph mining, and stream mining. The team develops methods for representing, integrating, and reasoning over complex, dynamic data to enable interpretable and trustworthy AI. Applications range from general-purpose AI to domain-specific areas such as healthcare and law. More specifically, the DIG team’s research activity covers the following topics:

Knowledge Graphs
Machine/Deep Learning
Language/Foundational Models
Natural Language Processing
Logic and Algorithms
Graph mining
Stream streams
Application Domains: Healthcare, Legal Domain

Industry Collaborations

Research Projects

Research

Algorithms:

AMIE: AMIE is a project to learn patterns and rules in ontologies. We conduct this project together with the Database department of the Max Planck Institute for Informatics in Germany.

Knowledge Graphs:

YAGO: YAGO is a large ontology constructed from WordNet, Wikipedia, and other sources. We develop YAGO together with the Database department of the Max Planck Institute for Informatics in Germany.

Softwares:

scikit-network: scikit-network is a Python package for the analysis of large graphs (clustering, embedding, classification, ranking).

We investigate how to do machine learning in real time, contributing to new open source tools:
- River: a Python library for online MachineLearning
- MOA: Massive Online Analytics, a framework for mining data streams (in Java)
- Apache SAMOA: Scalable Advanced Massive Online Analytics, an open source framework for data stream mining on the Hadoop Ecosystem

Language and Relevance

Computer science is not just about computers. In this area of research, we investigate how humans reason, and what this implies for machines.

Simplicity theory seeks to explain the relevance of situations or events to human minds.
Relevance in natural language: The point is to retro-engineer methods to achieve meaningful and relevant speech from our understanding of human performance.
We apply game theory and social simulation to explore conditions in which providing valuable (i.e. relevant) information is a profitable strategy. Read this paper.

Team

Talel Abdessalem	Mehwish Alam	Albert Bifet	Thomas Bonald	Jean-Louis Dessalles

Nils Holzenberger	Louis Jachiet	Van-Tam Nguyen	Nikola Simidjievski	Fabian Suchanek