TenderKG: A Knowledge Graph of French Public Tenders, 2021-2023

TenderKG is a domain-specific knowledge graph for public procurement, extracted from French public tender notices awarded from 2021 to 2023. It models the procurement ecosystem with entities like public buyers, contracts, lots, and bidding companies, enriched with CPV and NAF codes and geographical information. The dataset is provided in RDF format (TTL) and was created by Yacine Mokhtari.

Use Cases

• Developing bidder recommendation systems based on historical contract awards and company profiles.
• Knowledge graph learning and embedding research based on hierarchical CPV and NAF taxonomies.
• Studying market dynamics and competition in public procurement based on long-tailed distributions of company activity.
• Building graph machine learning models for cold-start and cross-domain generalization scenarios.

Strengths

• Integrates heterogeneous information from multiple public data sources, capturing key entities and relationships.
• Enriched with domain-specific side information including hierarchical CPV and NAF codes and geographical data.
• Models real-world characteristics like high sparsity and long-tailed distributions, providing a challenging benchmark.

Limitations

• Interactions are limited to awarded contracts, with no explicit negative or repeated interactions.
• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

Extracted from public notices of French public tenders by Intescia Group.

Collection Method

Result of Intescia Group's collection and processing pipeline.

Time Range

2021 to 2023

Freshness

Last updated 2026-05-03 22:05:10; freshness should be verified.

Geography

France