Entity Resolution Datasets Across Multiple Domains

This repository aggregates public entity resolution datasets from DeepMatcher, Magellan, and WDC sources. It contains labeled matching and non-matching entity pairs from relational tables across domains like product, citation, restaurant, and anime. The specific row count, column count, and dataset sizes are not provided.

Use Cases

• Train models to classify matching/non-matching entity pairs using labeled data from relational tables.
• Benchmark entity resolution algorithms across diverse domains such as product, citation, and restaurant.
• Analyze attribute-level similarity for entity pairs within specific domains like anime or books.

Strengths

• Aggregates datasets from three established public sources: DeepMatcher, Magellan, and WDC.
• Covers a variety of domains including product, citation, restaurant, and anime.

Limitations

• Key metadata such as row counts, column counts, and file sizes are unavailable.
• The sample data and specific file formats are not provided for inspection.

Provenance

Source

Public datasets from DeepMatcher, Magellan, and WDC.

Collection Method

Aggregated from existing public dataset repositories.

Freshness

Last updated on 2022-07-05.