Folksonomy meets rigorously defined common-sense
Integration of ontologies begins with establishing mappings between their concept entries. We map categories from the largest manually-built ontology, Cyc, onto Wikipedia articles describing corresponding concepts. Our method draws both on Wikipedia's rich but chaotic hyperlink structure and Cyc's carefully defined taxonomic and common-sense knowledge. On 9,333 manual alignments by one person, we achieve an F-measure of 90%; on 100 alignments by six human subjects the average agreement of the method with the subject is close to their agreement with each other. We cover 62.8% of Cyc categories relating to common-sense knowledge and discuss what further information might be added to Cyc given this substantial new alignment.
by Olena Medelyan and Catherine Legg, University of Waikato, NZ
Full
paper (to appear in Proceedings of the WikiAI Workshop at AAAI-2008,
Chicago, US)
Here you can download our mappings under the terms of Creative Commons license:
| Mappings | Description |
| 33,481 exact mappings | Highly accurate mappings. No disambiguation was required since only 1 possible mapping from Cyc to Wikipedia was available. |
| 52,690
(exact mappings plus semantically disambiguated ones) |
Fairly accurate mappings. Semantic disambiguation was applied. Wikipedia article with the highest similarity to the context of the given Cyc term was chosen, the commonness of the mapping was also taken into account. |
| 42,279
(exact mappings plus semantically disambiguated ones minus those that failed the common-sense test) |
The above mappings, after they have been corrected by removing some of the Cyc terms that have been mapped to the same Wikipedia article. |
The format of the mappings is the following:
Id of cyc term, name of cyc term, id of wikipedia article: title of wikipedia article, separated by tabs.
The ids correspond to Wikipedia version 2007/11/20.
Acknowledgements: We thank Mark Baltzegar, Sudarshan
Palliyil, Tim Wil-mot-Sitwell, Glen Meyer, Kendall Lister, Brett Summers,
Marios Daoutis and Cycorp Inc. for their generous help with this research.
,
2008