April 3, 2006 10:28 PM PDT
Pixsy unveils visual Web search
That's the virtual promise of Pixsy, a visual search engine that scours syndication feeds (in the format of Really Simple Syndication) for up-to-date images and then makes them searchable.
On Tuesday, the company will relaunch its engine with a revolving repository of millions of thumbnail images, which are drawn from photos and videos on sites ranging from The New York Times to YouTube.
"Anywhere there's an RSS feed, we consume it, extract an image...and make it searchable," said Chase Norlin, founder of the San Francisco-based company.
As opposed to search giant Google, which retrieves relevant pages from billions of Web sites, Pixsy hones in on the freshest images from publishers, Norlin said. "So you can now explore the Web visually."
For example, visitors can click a New York Times logo on Pixsy to see a collection of the newspaper's latest photos, which are then linked to news stories on the Times Web site. People can also type "George Clooney" in the search box to see photos of the Academy Award-winning actor, linked to all the latest stories about him.
The timing is apt. Multimedia is an increasingly large part of a reader's diet on the Web. Image search was the fastest-growing form of search on popular sites such as Google, Yahoo and MSN in the last year, up 91 percent from February 2005 to February 2006, according to a report from researcher Nielsen/NetRatings issued Monday.
Pixsy, a privately held company with four employees, officially launched its site last July. It had difficulty aggregating images, however, because the technology relied on XML feeds with a limited number of partners. (Norlin would not disclose the site's traffic.) The service now pulls images from hundreds of RSS feeds, according to Norlin, and that number is growing hourly, he said.
The site, which is built with AJAX technology, will collect a thumbnail image from an RSS feed automatically and then associate words, or metadata, with that image based on the news or information from where it came. Pixsy then uses that data to associate images with search terms.
Still, publishers could grumble over use of their images, even in the form of thumbnails. Pixsy does not have partnerships with publishers from which it draws, but rather relies on the inherent marketing push of RSS feeds. Norlin said he believes that publishers will be pleased with the traffic.
For now, Pixsy makes money through advertising and affiliate partnerships. For example, if an image seen on Pixsy drives traffic to a site that sells a poster of the image, it would collect a small fee. In the future, Norlin said, the company envisions licensing its visual search engine to other companies or publishers.
3 commentsJoin the conversation! Add your comment