Despite the continuing efforts of Microsoft, Yahoo! and others, Google remains the dominant horizontal search engine for most people, most of the time. In the United States, comScore reports 58.5% of searches during January were via a Google property. In the Semantic Web space, search is far less established and a number of much smaller sites offer their own solutions to the problem of locating appropriate semantic content from across the open web. Whether those sites are complementary or competitive seems to depend upon one's perspective, and it is also interesting to ponder the extent to which Yahoo!'s recent announcement is an attempt to position themselves as the search engine of choice for the growing web of semantically enriched content.
James Simmons published his Semantic Web Search Engine Roundup over on Semantic Focus a couple of weeks ago, and this does a useful job of pulling together some of the best known contenders in this area. I've also managed to have conversations with researchers involved with Watson, SWSE and Sindice, and my colleague Danny Ayers revisited Sindice in his recent chat with Richard Cyganiak.
Most of these, and the others that James mentions, are at an early stage in their development. They've also, on the whole, been built as part of wider research activities rather than as consumer-facing alternatives to the existing search engines. As with the interesting work that DBpedia is doing to extract machine-understandable meaning from the pages of Wikipedia, presentation and usability sometimes takes a back seat to technological pushing of boundaries. Which would you rather read? This from Wikipedia, or (more or less) the same from DBpedia? And which would you expect some software acting on your behalf to get more value from?
Yahoo!'s announcement last week may just have changed the game. Suddenly, the Semantic Web isn't just inside an application like Twine. It isn't just inside that small (but growing) group of data sets that the Linked Open Data project have reached. Suddenly, those little snippets of semantic value scattered all across the web are being watched and put to work by one of the big players, and we'll probably see a scramble by Google, Microsoft and Ask to take the covers off their own semantic skunk works; a scramble matched by that on the other side of the divide as search engine optimisation companies rush to work out how to milk this latest opportunity.
The world in which Sindice, Swoogle, Watson et al have been quietly working just got noticed. Will they seize the opportunity, or be trampled in the stampede?