Last year, when Wikipedia founder Jimmy Wales announced plans to launch a new search engine in the first half of 2007, everyday users of this now ubiquitous tool wondered what Wales could do that Google couldn’t.
However, the search engine community knew better. Google may be an official verb in the dictionary, but the conventional wisdom is that the search engine giant can’t dominate the field indefinitely.
The Searcher’s Intention
Universities, startups and even established players such as MSN are experimenting with new technologies in the hopes of establishing a foothold in the next generation of search.
As discussed in Part 1 of this two-part series, many of these efforts are aimed at discerning the intention of the searcher.
“That is where the next battle will be,” Debby Richman, senior vice president of search engine provider Collarity, told TechNewsWorld. “Search is just in the kindergarten stage now.”
Industry by Industry
These firms are also approaching search technology from a vertical or industry-specific approach.
Microsoft, for instance, recently acquired the healthcare search engine Medstory with an eye to expanding its footprint in the consumer healthcare industry. Other companies are trying to build similar functionality from the ground up.
“Verticalization is one of the search engine trends that is just emerging now but [is] already having a significant impact on direction and planning,” Chris Weiss, a product management with search technologies firm Convera, told TechNewsWorld.
The benefits to this approach are obvious: A consumer looking for information on breast cancer is more likely to find what she is looking for in a medical search engine than a general search, which could pull in links to everything from fast food to clothing stores.
“Any industry from travel to games to medical can benefit,” Weiss said.
Some companies are taking this concept a step further, linking search engines to communities that can provide feedback on the subject matter.
“Integrating the current search algorithms with social networks is already proving to be very popular,” Garrett Camp, cofounder and chief architect of Stumbleupon, told TechNewsWorld.
For instance, someone looking for wine in a general search engine will pull up a myriad number of links. A specialized search engine could direct the searcher to wines from California’s Napa Valley far more easily. If that engine is overlayed on a network, other people in the community can help rank or rate the links, he noted.
Image search is also gaining momentum among companies interested in developing next generation search technology. Of course, there are a number of commercially available image search engines on the market today, P. Kevin Smith, vice president of sales, North America, for LTU Technologies, told TechNewsWorld.
These mostly use “tags” or keywords associated with images to ultimately find matching images. However, this approach has proven to be unreliable, he stated.
“A common issue with this approach is incomplete or less than thorough search results,” Smith noted, adding, “Likewise, it’s very common for users to complain that ‘tag-based’ search ultimately does not accurately match or find significant numbers of relevant images known to be in their reference sets, such as databases or repositories.”
Advanced image search tools will likely incorporate the use of pixel or content-based image search, he explained. Consumers will experience far more accurate results when using technologies that incorporate this “content-based” methodology, Smith noted.
Currently in Use
This technology is already being used by law enforcement agencies as well as some patent offices.
One of the products LTU offers, for instance is Image-Filter, a Web-based tool used to detect specific types of image content in real time. A police or federal agent can use it to identify pornographic content that is attached to e-mails, being uploaded to social networking sites or shared in IM or chat sessions.
Corporate security is another use, to prevent employees from downloading or e-mailing proprietary information.
“Over time, this product could be made available as a desktop application sold directly to consumers,” Smith speculated. School systems and public libraries are another possible venue for such applications.
Retailers Rush to Adopt
Forms of advanced image search are already used on some retail Web sites, Erik Hansen, president of search engine marketing provider SiteSpect, told TechNewsWorld.
In theory, a retail image search engine would allow a searcher to look for, say, a piece of jewelry she saw an actor wearing during the Oscars. The searcher can plug in that image, and then search for jewelry similar to that piece.
One Web site, www.Like.com, is using a variation of this approach to drive sales, Hansen noted. Searchers can click on a picture of a purse, for example, to see similar products at different price points, by other designers and in other colors.
Retailers are likely to be among the earliest adopters of next gen image search that comes on the market, he stated.
“For them, it is all about driving traffic to their site and then keeping up the momentum,” Hansen concluded.
It is the same concept behind the current crop of search engines — just a vastly different execution.