Google Pictures Better Search Results

Google is proving that a picture can say a thousand words, or in its case, 1.1 billion pictures can say enough to fill a library.

Today, Google announced that its image search property's index has swelled to over 1.1 billion pictures from sources worldwide. Before long, the images themselves will be displayed against the traditional backdrop of aggregated search results contributing what Google refers to as "Image Results."

Once Images Results are integrated into Google's main Web search, queries such as "sun sets" will invoke thumbnail images that are linked directly to the image library. Thumbnails will be placed prominently on top of search results.

Unlike search rival AOL, which produces its own image results called Snapshots, Google Image Results lack structured editorial content. Image Results are more comparable to Amazon's A9 Yellow Pages, which displays pictures of adjacent storefronts for local phone numbers or addresses.

The emergence of pictures in Web searches reflects a wider trend, as companies eschew the customary list of links in favor of better, more organized results. America Online is a prime example of the practice; making search results easier to navigate is the hallmark of its strategy.

Last month, AOL re-launched AOL Search to include expanded Snapshot categories, as well as a compilation of locally relevant information. AOL Search also clusters related content, categorizing results by deciphering relationships.

Instead of earmarking resources for the development of search functionality, AOL licenses technology from Fast Search & Transfer (FAST) and Vivismo so that it can focus exclusively on presentation and usability.

Google Labs, however, has developed competing in-house technologies that blaze the same trail as AOL and Amazon. Although AOL's search enhancements have already gone live, Google is not resting on its laurels. In addition to Images Results, Google is continuing the ongoing development of a local search engine and "named entities abstraction" clustering.