Thursday, May 17, 2007

Google Wants The Best Answer


"Back in 2001, Eric asked for a brainstorm of a few "splashy" ideas in search," says Marissa Mayer, Google VP. "I made a few mockups, one of which was for "universal search," she says. "It was a sample search results page for Britney Spears that, in addition to web results, also had news, images, and groups results right on the same page."

"Even then, we could see that people could easily become overwhelmed with the number of different search tools available on Google, let alone those that would be created over the next few years," Meyer says. "This proliferation of tools, while useful, has outgrown the old model of search," says Meyer. "We want to help you find the very best answer, even if you don't know where to look."

That mockup and early observations were the motivation behind the universal search effort Google expects will break down silos of information that exist on the Web. Google's vision for universal search is to ultimately search across all its content sources, compare and rank all the information in real time, and deliver a single, integrated set of search results that offers users precisely what they are looking for. Google already is incorporating information from a variety of previously separate sources – including videos, images, news, maps, books, and websites – into a single set of results.

In principle, the new approach means that searches on a single key word or phrase will return unified results that might previously have required separate searches in "Web," "news," "video," "image," "people" and "maps" engines.

Google also is deploying a new technical infrastructure that will enable the search engine to handle the computationally intensive tasks required to produce universal search results and releasing the first stage of an upgraded ranking mechanism that automatically and objectively compares different types of information.

New dynamically generated navigation links have been added above the search results to suggest additional information that is relevant to a user's query. For example, a search for "python" will now generate links to Google Blog Search™, Google Book Search™, Google Groups™, and Google Code™, to let the user know there is additional information on his or her query in each of those areas.

As a result, users can find a wider array of information on their topic, including data types they might not have initially considered.

Google's homepage and a number of applications have also been updated with a new navigation bar to provide easier access to popular Google products. Now, instead of having links above the Google.com homepage search box, users will see a navigation bar on the top left side of the page with various Google search properties and popular products including Gmail™, Google Calendar™, Google Docs & Spreadsheets, and Picasa Web Albums.

Google also announced a new experimental version of its popular search service called Google Experimental, available on Google Labs. This new test site provides users an opportunity to try out some of the latest search experiments and innovations and provide Google with feedback.

One of the first experiments to be featured on the site enables users to view their search results on a map or timeline. For instance, when someone searches for "Albert Einstein" on Google Experimental, they can choose to view the search results on a map that shows locations mentioned within web pages about Albert Einstein or on a timeline that illustrates the history of Albert Einstein's life.

No comments:

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...