Sylloge: Unified Loader for Entity Alignment Benchmarks

Sylloge is a collection and loader library for entity alignment benchmark datasets, maintained by developer dobraczka and last updated in June 2024. It provides a unified interface for accessing multiple knowledge graph datasets specifically curated for entity resolution and alignment tasks.

Use Cases

• Benchmarking entity alignment algorithms using provided knowledge graph triples
• Evaluating entity resolution models on cross-source entity pairs
• Testing knowledge graph embedding techniques on standardized benchmark sets

Strengths

• MIT license allows for unrestricted academic and commercial use
• Specialized focus on entity alignment and resolution benchmarks
• Standardizes the loading process for multiple disparate knowledge graph sources

Limitations

• Specific record counts and schema details are not provided in the primary metadata
• Functionality is dependent on the availability and persistence of external benchmark sources

Provenance

Source

GitHub repository by user dobraczka

Collection Method

Aggregated from existing entity alignment benchmark datasets

Freshness

Last updated June 2024; frequency of new benchmark additions is not specified.