Research

Significant challenges in artificial intelligence and data management arise from the inherent characteristics of data. Data can vary widely in structure, size, dimensionality, and quality—it may be semi-structured, large, high-dimensional, incomplete, or noisy. On the other hand, data can also follow well-defined logical rules, with inherent patterns and constraints that can be leveraged to improve processing, inference, and decision-making.

In the Tyrex project-team, we advance research at the interface of data management and artificial intelligence.

Projects

GraphRec (2024-..)

Through the ANR-funded GraphRec project, we investigate recursive navigational graph query processing, relational learning, and neuro-symbolic approaches to querying large graphs. Recursive navigational graph queries enable the exploration of paths of arbitrary length within graph-structured data, allowing efficient retrieval of relationships that are not explicitly stored but can be inferred through recursion. This is particularly useful for applications such as social network analysis and knowledge graph reasoning. Additionally, we study relational learning techniques to extract meaningful patterns from graph data and integrate symbolic reasoning with neural methods to enhance query efficiency and interpretability.

CLEAR (2017-2022)

Through the ANR-funded CLEAR project, we explored programming techniques to enhance the extraction of value from large-scale data. Our work included developing efficient methods for querying large graphs, generating optimized distributed code (e.g., for Spark) to process massive datasets, and designing scalable predictive models.