Google’s attempts to innovate their search results have taken an interesting turn. According to the official Google blog, the search giant is striving to create a more intelligent engine that puts users in front of “things not strings”. In other words, Google is expanding their focus on the world of linked data and the semantic web to try and present meaningful relationships between individual data points.
Amit Singhal suggests that search engines have, up to this point, treated queries as just strings of words and this isn’t good enough. Human beings know that terms such as [taj mahal] and [van gogh] have richer meanings than being just two words strung together. Google’s long term goal is to show users a results page that interprets and displays those richer meanings. They call this Google’s Knowledge Graph.
On the linked data side of things, Kingsley Idehen from OpenLink Software posted a theory about how the Google Knowledge Graph works. He suggests that Google is pulling information from users’ profiles and using it to present different results to people with different backgrounds.
Mr. Idehen’s suggestion is interesting because it would mean that Google has taken personalized search to a new level. Rather than operating off of a basic PageRank model where the number of backlinks and, now, the number of +1′s and social signals impact results, they are using structured data sets combined with individual situations to present results. In simpler terms, it’s a more sophisticated model than just counting signals.
Mr. Idehen argues that the new changes seem to have been rolled out to journalists, analysts, and bloggers on a limited basis. I personally don’t see Google’s knowledge graph yet, so I’m afraid I have to miss out on the fun.
I’ve collected a list of interesting and helpful Facebook tips. There are no shortages of articles that promise tips for Facebook, but there is a lack of organization for good articles.
