You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am also interested in whether the graph-based information content similarity metrics can be applied to other networks than
DBpedia.
For example, I'd like to compute Resnik or Lin similarities on the Experimental Factor Ontology. Is this library general purpose or specialized just for work for DBpedia?
Since the original author is not working at GSI anymore, allow me to answer your question to the best of my ability.
For the most part, sematch was conceived as a general purpose tool. There are some parts that seem to be very specific to DBpedia (e.g., the NameSPARQL class), but most of them could be used with any other SPARQL endpoints / RDF graphs despite their name.
I'm not really familiar with the Experimental Factor Ontology, but my guess is that you could create an ad-hoc DataTransform, or reuse the existing DBpediaDataTransform pointing it to EFO's .owl file. That should be enough to calculate concept similarity. You can check the documentation and the test_similarity.py.
We will be doing maintenance work on sematch, and we might include other data sources if the community shows enough interest.
Can the tool be used to get semantic similarity between the entities of other knowledge bases such as conceptnet.
The text was updated successfully, but these errors were encountered: