Benchmark Datasets for Schema and Ontology Matching
Benchmark Datasets for Schema and Ontology Matching
Datasets
The table below comprises a set of selected match tasks that have been solved with COMA++. Each match task consists of at least two source schemas (owl or xdr) to be matched against each other. We also provide the presumably perfect mapping result that can be used for evaluating the results of automatic match approaches. If more than two schemas were matched, the mapping file contains the mappings for all possible schema combinations.
Please refer to the papers to see how we determined the perfect mapping.
Domain | Type | #Elements | #Corresp. | #Src | Instances? | Sources | Used in |
---|
Purchase Order (Apertum, CIDXPOSCHEMA, Excel, Noris, Paragon) | xdr | 35 .. 82 | 36 .. 85 | 5 | No | Purchase Order | [1], [2], [3], [4] | Web Directory (Google, Yahoo, web, dmoz) | owl | 418 .. 1132 | 197 .. 792 | 4 | Yes | WebDirectory | [3], [4], [5] | Lebensmittel (Google, web) | owl | 53 .. 59 | 32 | 2 | Yes | Lebensmittel | [3] | Freizeit (Google, dmoz) | owl | 67 .. 71 | 67 | 2 | Yes | Freizeit | [3] |
Links to collections of further match tasks
Illinois Semantic Integration Archive
Trento tasks
The Open Biological and Biomedical Ontologies (OBO Foundry)
Ontology Alignment Evaluation Initiative
Test cases for Entity Resolution
Publications
Contact/Project Members
|