Swoogle is experimenting with a new interface that searches for semantic web documents using terms whose local names contain certain substrings. This seems to provide a good way to find documents about a given topic. For example, to find ontologies about time, you might search for documents using terms matching before after time instant. For efficiency reasons, we are not matching on a substring, but rather decompose the local name into one or more lexemes. For example, the local name BeliefConnective is decomposed into {belief connective}. You can, of course, still search using substrings, but queries contaning such constraints can be expensive.

Related posts:

1. MSN splitting the cake of map/local search
2. Finding RDF instance data with Swoogle
3. Semantic Web Word Search
4. Swoogle namespace searches
5. Swoogle: over 1,000,000 Semantic Web documents