BIT-DataLab/LakeBench resource

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1211 stars · 36 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

LakeBench is a large-scale benchmark for evaluating table discovery methods (joinable and unionable table search) in data lakes, containing over 16 million real tables and 10 thousand queries. It includes implementations of state-of-the-art join and union search algorithms for comparison.

Use cases

• benchmark table discovery algorithms on large data lakes
• evaluate joinable table search methods
• evaluate unionable table search methods
• compare state-of-the-art table search systems like DeepJoin and Starmie
• research scalable schema matching in data lakes

When to choose

• you need a realistic large-scale benchmark for table discovery research
• you want to evaluate join or union search algorithms at scale
• you need reproducible baselines for data lake table search

When to avoid

• you need a production table discovery system rather than a benchmark
• your data lake is small and existing smaller benchmarks suffice
• you need a non-Python or non-PyTorch evaluation environment

Facets

dataset · maturity active

benchmarking search-engine data-science machine-learning databases data-science big-data analytics python cross-platform table-discovery data-lakes join-search union-search benchmark-dataset

Member repositories

For agents

markdown · JSON · MCP: product_card(name="BIT-DataLab/LakeBench")