CoolFace
Datasetpublic

Urwashanza/europrocure-10-public-procurement

EUROPROCURE-10 Dataset Summary EUROPROCURE-10 is a large-scale, cleaned, enriched, and machine-learning-ready dataset of European public procurement notices derived from TED (Tenders Electronic Daily), the official procurement platform of the European Union. The dataset covers procurement activity published between 2016 and 2025 and contains structured procurement metadata, financial information, supplier information, CPV classifications, procurement procedures… See the full description on the dataset page: https://huggingface.co/datasets/Urwashanza/europrocure-10-public-procurement.

sourceHugging Faceupdated 3mo agoView on Hugging Face
0likes20downloads
Dataset Card

EUROPROCURE-10

Dataset Summary

EUROPROCURE-10 is a large-scale, cleaned, enriched, and machine-learning-ready dataset of European public procurement notices derived from TED (Tenders Electronic Daily), the official procurement platform of the European Union.

The dataset covers procurement activity published between 2016 and 2025 and contains structured procurement metadata, financial information, supplier information, CPV classifications, procurement procedures, and AI-enriched features designed for analytics and machine learning applications.

Key Statistics

MetricValue
Records1,500,356
Columns51
Coverage Period2016–2025
Countries & Territories190
Languages36
Procurement Domains7
Unique Buyers132,849
Unique Winning Suppliers216,228

Supported Tasks

This dataset can be used for:

  • —Procurement analytics
  • —Supplier intelligence
  • —Public-sector spending analysis
  • —Fraud-risk detection
  • —Weak supervision research
  • —Anomaly detection
  • —Procurement NLP
  • —Information extraction
  • —Classification
  • —Forecasting
  • —Market intelligence

Procurement Domains

Each notice is assigned to one of seven procurement domains:

  1. 1.Healthcare
  2. 2.Digital Technology
  3. 3.Construction & Infrastructure
  4. 4.Automotive
  5. 5.Green Energy
  6. 6.Defense & Security
  7. 7.Aerospace

Dataset Structure

Main Categories

Procurement Metadata
  • —publication_number
  • —publication_date
  • —publication_year
  • —publication_month
  • —notice_type
  • —notice_stage
Buyer Information
  • —buyer_name
  • —buyer_country
  • —buyer_city
  • —nuts_region
  • —contact_email
  • —contact_phone
Supplier Information
  • —winner_company
Financial Information
  • —estimated_value
  • —award_value
  • —currency
  • —awardestimateratio
  • —value_tier
  • —awardvaluelog
  • —value_disclosed
Classification Data
  • —cpv_codes
  • —cpvprimarycode
  • —cpvsecondarycodes
  • —cpv_description
Procedure Information
  • —procedure_type
  • —contract_type
  • —numberofbids
  • —submission_deadline
  • —daystodeadline
Text Features
  • —tender_description
  • —technical_requirements
  • —eligibility_requirements
  • —award_criteria
AI-Enriched Features
  • —keywords
  • —aigeneratedtags
  • —procurementrisklevel

Data Processing

The final release includes:

  • —Removal of redundant technical fields
  • —Notice-type normalization
  • —Country-code normalization
  • —Language-code normalization
  • —CPV code cleanup
  • —Procurement-domain classification
  • —Keyword extraction
  • —AI-generated semantic tags
  • —Procurement risk-level generation

Original TED source information was preserved wherever possible while improving analytical consistency.

Missing Data

Several fields contain missing values due to procurement-stage differences and reporting practices.

ColumnMissing %
award_value56.1%
estimated_value50.2%
winner_company58.7%
numberofbids65.2%
nuts_region70.8%
eligibility_requirements73.6%
eufundname94.5%

These missing values generally reflect source reporting limitations rather than extraction failures.

Financial Data Warning

awardvalue and estimatedvalue are stored in their original notice currencies.

The dataset contains multiple currencies and values have not been converted to EUR.

Users should not aggregate or compare financial values across currencies without applying appropriate exchange-rate normalization.

Additionally, some framework agreements report maximum spending ceilings rather than realized expenditures.

Source Data

Raw procurement data was sourced from TED (Tenders Electronic Daily), operated by the Publications Office of the European Union.

TED website:

https://ted.europa.eu/

License

Raw TED data is available under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.

License:

https://creativecommons.org/licenses/by/4.0/

Attribution

Raw data sourced from TED (Tenders Electronic Daily), European Union.

Enriched and processed by Urwa Binat Khalid.

Additional enrichments include:

  • —Procurement domain classification
  • —Keyword extraction
  • —AI-generated tags
  • —Procurement risk indicators
  • —Feature engineering
  • —Standardized procurement metadata

Disclaimer

The European Union, TED, and the Publications Office of the European Union do not endorse this dataset, its derived attributes, analyses, conclusions, machine learning models, visualizations, or associated products.

All enrichments, classifications, and derived fields were independently produced by the dataset author.

Citation

If you use EUROPROCURE-10 in research, publications, or commercial applications, please cite:Zenodo DOI: 10.5281/zenodo.20690042

Raw data sourced from TED (Tenders Electronic Daily), European Union. Enriched and processed by Urwa Binat Khalid.