BigShinyThing

Search [engines] will ultimately be as good as having 1,000 human experts who know your tastes scanning billions of documents within a split second. It will model the human brain. - Gary Flake, one of the seven Distinguished Engineers at Microsoft.

This week, Time Magazine writes about the future of search. According to the article, the battle is on for the next generation of search - one that can successfully harvest the projected $22bn worldwide that search-engine advertising is expected to be worth by 2010.

Google remains the market leader in search handling 36.5% of queries. Yahoo! is surprisingly close behind, with 30.5% and MSN has 15.5%, according to comScore Media Metrix. The Big Three, as they are termed, are investing aggressively in search technology - Google alone earmarked $4bn earlier this year for growth. But Time identifies a group of start ups, all of course acquisition targets for the Big Three, that may also hold the key to the future of search.

Some of them, like photo sharing site Flickr, have already been pounced on - it was bought by Yahoo! earlier this year. Yahoo! was responding quickly to the growing phenomena of tagging - a simple device whereby users label websites with descriptive tags, building a folksonomy- a taxonomy of knowledge organised by ordinary people. Flickr is organised with a communal tagging model and Yahoo! have already developed a tagline to market tags across all of its user-created content: “Better search through people.”

Open source stuff like Flickr is already being used to enhance functionality for community sites. This week street art collective Wooster launched Streetsy, essentially a graffiti photo library which is on Flickr and is catalogued using its tags (see BST article ‘Coming up from the streets’), . The commercial world has also become wise to the potential of enhanced search: ABC now invites users to view news locations via Google Earth. The news site knows a competitor when it sees one: Google Earth is currently compiling an overview of Hurricane Katrina’s devastation - it is likely to be the definitive source of images of the catastophe.

Meanwhile, A9.com, Amazon’s search subsidiary has sent trucks around 22 US cities with digital cameras linked to laptops to photograph every street. So far it has 35 million pictures, which will be overlaid on maps. Microsoft is combining the approaches from the air - its Virtual Earth project is flying planes over cities to take pictures. The aim is to have views from all directions so users can circle buildings onscreen. The aim, according to MSN, is to create “a fully immersive virtual-reality experience.”

BST has already covered one of the most touchy feely new search facilities: Blinkx.TV which can find video clips based on dialogue searches. Speech recognition technology is now improving so rapidly that the company can capture the audio tracks of videos and turn them into searchable text - making any recorded spoken word searchable.

The next step for plain old word search is semantic search - looking for meaning, not just text. Oren Etzioni, a University of Washington computer scientist, uses language-analysis programs to power KnowItAll, which scans documents for facts. So far, KnowItAll has extracted 900 million facts - enabling it to answer questions. Nosa Omoigui, a former Microsoft researcher, has founded Bellevue-based Nervana, which analyses language by linking word patterns contextually to answer questions in defined subject areas, such as medical research literature. In short, the net is soon going to be able to read and research for us.

The marketing press also reported this week that Google has moved into offline advertising, buying up print ad space in tech titles such as PC Magazine and Maximum PC. The print ads are labelled as “Ads by Google.” Joshua Stylman, managing partner at search marketing firm Reprise Media, says of the move: “Inevitable is my gut reaction. Search marketing works because of the whole relevancy factor. The model is transcendent, and clearly, the model is scalable. It was a matter of time for it to start to migrate offline.”

Maybe the vision of an utterly converged future - where Amazon, Google and Microsoft become the all powerful ‘Googlezon‘ and swallow all offline media - isn’t that far off.