GodotCN/science-datalake

Science Data Lake

A unified, portable science data lake integrating 7 scholarly datasets (~525 GB Parquet) with cross-dataset DOI normalization, 13 scientific ontologies (1.3M terms), and a reproducible ETL pipeline.

Note: One additional source (Semantic Scholar S2AG) is supported by the pipeline but is not redistributed here due to its API terms of service. See Not Included in This Upload below.

What's Unique

This dataset enables queries that are impossible with any single source:

-- "Top disruptive papers with open-source code, checking for retractions"
SELECT doi, title, year,
       sciscinet_disruption,      -- from SciSciNet
       oa_cited_by_count,         -- from OpenAlex
       has_pwc,                   -- from Papers With Code
       has_retraction             -- from Retraction Watch
FROM unified_papers
WHERE has_pwc AND sciscinet_disruption > 0.5
ORDER BY oa_cited_by_count DESC
LIMIT 20

Datasets Included

Dataset	Papers/Records	License	Key Contribution
OpenAlex	479M works	CC0 1.0 (public domain)	Broadest coverage, topics, FWCI
SciSciNet v2	250M papers	CC BY 4.0	Disruption index, atypicality, team size
Papers With Code	513K papers	CC BY-SA 4.0	Method-task-dataset-code links
Retraction Watch	69K records	Open (via Crossref)	Retraction flags + reasons
Reliance on Science	47.8M pairs	CC BY-NC 4.0	Patent-to-paper citation pairs (global)
Preprint-to-Paper	146K pairs	CC BY 4.0	bioRxiv preprint to published paper
13 Ontologies	1.3M terms	Various (see below)	CSO, MeSH, GO, DOID, ChEBI, NCIT, HPO, EDAM, AGROVOC, UNESCO, STW, MSC2020, PhySH

Ontology Licenses

Ontology	License
MeSH	Public Domain (US government work)
GO, ChEBI, NCIT, EDAM, CSO, PhySH, STW	CC BY 4.0
DOID	CC0 1.0
AGROVOC	CC BY 3.0 IGO
UNESCO Thesaurus	CC BY-SA 3.0 IGO
HPO	Custom (free for research use)
MSC2020	CC BY-NC-SA 4.0 (non-commercial)

Snapshot Dates

Each source was downloaded at a specific point in time:

Dataset	Snapshot / Release	Notes
OpenAlex	2026-02-03	S3 snapshot
SciSciNet v2	2024-11-01	GCS bucket
Papers With Code	2025-07	Archived JSON
Retraction Watch	2025-02	Crossref CSV
Reliance on Science	v64	Zenodo record
Preprint-to-Paper	2025-06	Zenodo record
13 Ontologies	2026-02	Official sources

All snapshots can be refreshed using the update pipeline — see below.

Not Included in This Upload

The following source is supported by the full pipeline (GitHub) but is not redistributed here due to its API terms of service:

Dataset	Reason	How to obtain
S2AG (Semantic Scholar, 231M papers)	License requires individual agreement with Semantic Scholar	Semantic Scholar Datasets API

After downloading S2AG locally, run the full pipeline to integrate it.

Key Tables

unified_papers (293M rows)

The headline table: one row per unique DOI, joining all sources.

Column	Type	Description
doi	VARCHAR	Normalized DOI (lowercase, no prefix)
title	VARCHAR	Best available title (OpenAlex > S2AG)
year	BIGINT	Publication year
openalex_id	VARCHAR	OpenAlex work ID
sciscinet_paperid	VARCHAR	SciSciNet paper ID
has_openalex	BOOLEAN	Present in OpenAlex
has_sciscinet	BOOLEAN	Present in SciSciNet
has_pwc	BOOLEAN	Has code on Papers With Code
has_retraction	BOOLEAN	Flagged in Retraction Watch
has_s2ag	BOOLEAN	Present in Semantic Scholar
has_patent	BOOLEAN	Cited by at least one patent (RoS)
s2ag_corpusid	BIGINT	Semantic Scholar corpus ID
s2ag_citationcount	INTEGER	S2AG citation count
oa_cited_by_count	BIGINT	OpenAlex citation count
sciscinet_disruption	DOUBLE	Disruption index (CD index)
sciscinet_atypicality	DOUBLE	Atypicality score
oa_fwci	DOUBLE	Field-Weighted Citation Impact

Note: The S2AG columns (s2ag_corpusid, s2ag_citationcount, s2ag_influentialcitationcount, s2ag_isopenaccess, has_s2ag) are present in the uploaded file but will contain NULL/FALSE values unless S2AG has been integrated locally. All other columns (including has_patent from Reliance on Science) are fully populated.

topic_ontology_map

Maps OpenAlex's 4,516 topics to terms in 13 scientific ontologies via embedding-based semantic similarity (BGE-large-en-v1.5, 1024-dim) + exact matching for large ontologies (MeSH, ChEBI, NCIT). 16,150 mappings covering 99.8% of topics. Columns include similarity (cosine, 0-1) and match_type (label/synonym/exact) for quality filtering.

ontology_bridges

Cross-ontology links discovered via shared external IDs (UMLS, Wikidata, MESH, etc.).

Usage with DuckDB

Option 1: Pre-built database file (recommended)

This repository includes a ready-to-use DuckDB database file (datalake.duckdb, 274 KB) with 145 SQL views pre-configured to read directly from HuggingFace. Download just this one file and query all 7 datasets immediately — no pipeline setup required.

import duckdb

con = duckdb.connect()
con.execute("INSTALL httpfs; LOAD httpfs;")
con.execute("ATTACH 'hf://datasets/J0nasW/science-datalake/datalake.duckdb' AS lake")

# Query using familiar schema.table syntax
df = con.execute("""
    SELECT doi, title, year, sciscinet_disruption, oa_cited_by_count
    FROM lake.xref.unified_papers
    WHERE sciscinet_disruption IS NOT NULL
    ORDER BY sciscinet_disruption DESC
    LIMIT 100
""").df()

# Cross-source joins work out of the box
con.execute("""
    SELECT t.display_name AS topic, o.ontology, o.term_name, o.similarity
    FROM lake.xref.topic_ontology_map o
    JOIN lake.openalex.topics t ON t.id = o.topic_id
    WHERE o.similarity >= 0.85
    ORDER BY o.similarity DESC
    LIMIT 20
""").df()

Option 2: Direct Parquet queries

You can also query individual Parquet files directly without the database file:

import duckdb

con = duckdb.connect()
con.execute("INSTALL httpfs; LOAD httpfs;")

df = con.execute("""
    SELECT doi, title, year, sciscinet_disruption, oa_cited_by_count
    FROM 'hf://datasets/J0nasW/science-datalake/xref/unified_papers/*.parquet'
    WHERE sciscinet_disruption IS NOT NULL
    ORDER BY sciscinet_disruption DESC
    LIMIT 100
""").df()

Keeping the Data Current

The full pipeline supports incremental updates. When upstream sources release new snapshots:

# Update a single dataset
python scripts/datalake_cli.py update openalex

# Update all datasets and rebuild cross-reference tables
python scripts/datalake_cli.py update
python scripts/materialize_unified_papers.py

See the GitHub repository for full pipeline documentation.

LLM & AI Agent Integration

This data lake ships with SCHEMA.md — a structured reference file optimized for LLM-based coding agents (Claude Code, Cursor, Copilot, etc.). It contains every table, column, type, join strategy, and performance tier in a format that AI agents can use to write correct DuckDB SQL without prior schema knowledge.

Point your AI assistant at SCHEMA.md and ask it to query across all 7 hosted datasets and 13 ontologies using natural language.

Building the Full Instance (All 8 Sources)

Clone the GitHub repository and run the pipeline to integrate all sources including S2AG:

git clone https://github.com/J0nasW/science-datalake
cd science-datalake
python scripts/datalake_cli.py download --all
python scripts/datalake_cli.py convert --all
python scripts/create_unified_db.py
python scripts/materialize_unified_papers.py

Citation

If you use the Science Data Lake, please cite the paper:

@article{wilinski2026sciencedatalake,
  title   = {The Science Data Lake: A Unified Open Infrastructure Integrating
             293 Million Papers Across Eight Scholarly Sources with
             Embedding-Based Ontology Alignment},
  author  = {Wilinski, Jonas},
  journal = {arXiv preprint arXiv:2603.03126},
  year    = {2026},
  url     = {https://arxiv.org/abs/2603.03126}
}

Dataset DOI: 10.57967/hf/7850

License

This dataset aggregates multiple sources, each with its own license. Users must comply with the most restrictive license applicable to the sources they use.

Component	License
Integration code (scripts, pipeline)	MIT
OpenAlex data	CC0 1.0 (public domain)
SciSciNet v2 data	CC BY 4.0
Papers With Code data	CC BY-SA 4.0
Retraction Watch data	Open (via Crossref)
Reliance on Science data	CC BY-NC 4.0
Preprint-to-Paper data	CC BY 4.0
Cross-reference tables (unified_papers, topic_ontology_map)	Derived work — most restrictive source license applies
Ontologies	Various — see table above; note MSC2020 is CC BY-NC-SA 4.0