X

Google Knowledge Graph hands-on: The Web's tail is long indeed

Hitting entries in Google's 500-million-item database of things is still a bit of a roulette spin.

Rafe Needleman Former Editor at Large
Rafe Needleman reviews mobile apps and products for fun, and picks startups apart when he gets bored. He has evaluated thousands of new companies, most of which have since gone out of business.
Rafe Needleman
2 min read
The Knowledge Graph returns great information (right of screen) on searches for popular media and other very mainstream items. It's not quite so good for the "long tail." Screenshot by Rafe Needleman/CNET

I got the magic cookie to turn on the new Google Search Knowledge Graph feature yesterday.

My searches now ping Google's 500-million-strong database of "things." So in addition to standard list-of-links search results, I get sweet data, beautifully packaged -- when there's a Knowledge Graph hit -- in place of an empty sidebar or one filled with ads.

Except for when I don't get the Knowledge Graph, which is most of the time.

It turns out that searching for "things" isn't something I do a lot of the time. If I were busy writing an essay on the Empire State Building, I would end up getting more Knowledge Graph hits. Or if I spent more time looking for famous people. Wikipedia searches, in other words.

If you're searching for particular things or people, Wikipedia is actually more likely to give you results than Google. The Knowledge Graph doesn't have everything Wikipedia does. Products and companies, for example, are mostly absent from the Graph, even when they exist in Wikipedia.

Search for a non-proper noun, like "trebuchet," and nothing comes up. Search for a concept, like "mortgage," where an explanatory dossier and list of largest mortgage banks would be illustrative, and you get nothing from the Knowledge Graph. Look for a news item ("Facebook IPO"), and also nothing. There's not even an entry for "Facebook" itself. Nor "Amazon," although there is one for Apple. (It lists related people, but no corporate details.)

When you do hit on a rich ore of Knowledge Graph data, it's very useful. Search for information on a mainstream movie or TV show, and you'll find yourself swimming in useful cross-links of directors, actors, and similar shows or films.

In my own use of Google for the past 24 hours, I have activated the Knowledge Graph only a few times. The 500 million items in the Graph database feel like not very many. "Search problems are very long tail," Jack Menzel, Google's product management director for search, told me.

So for now, hitting a Knowledge Graph results, while delightful, is rare. That's not necessarily bad, but it does show how personal search is, and how difficult it is to categorize the questions that people have for the collection of knowledge stored on the Web. Five hundred million pieces of knowledge, it turns out, is nothing compared with human curiosity.