Profiling Could Improve Enterprise Search

Ann All

Despite a number of developments in the enterprise search space this year by vendors, including Microsoft's moves to offer products across a broad spectrum of price and sophistication and Google's recent retooling of its Google Mini search appliance, many business users still find their search experiences seriously wanting.


According to a study by research group AIM, 49 percent of business users do not like their company's search software and find locating information difficult and time-consuming. Also, reports, 69 percent of respondents say less than half of their company's information can be accessed by online search tools and 38 percent estimate that search queries generate less than a quarter of the information they need.


Interestingly, rather than faulting available search technology, AIM researchers blame companies for not taking search more seriously, noting that few companies have specific plans or formal goals for collecting and indexing information. Just 10 percent of the businesses queried in the survey considered the ability to find documents "imperative."


A Computerworld article does a nice job of capturing the current state of the enteprise search market, laying out its potential opportunities and challenges.


Analyst Stephen Arnold, who authored a 300-page study of the market for Gilbane Group, identifies a number of types of search systems, including those focused on content that resides in databases, tools that help companies organize and prepare content to be searched, and systems employing "deep analysis" of content that will make keyword indexing look like a "Model-T Ford." Arnold taps Google as the company to watch.


Indeed, another Computerworld article details a new approach apparently being developed by Google that centers around an idea called "dataspaces." As explained in a Google research paper, dataspaces should contain all of the information relevant to a particular organization regardless of format and location, and the ability to illustrate the relationships between various data repositories. Though Google isn't the only vendor pursuing this type of approach, its scale gives it a key advantage, says Arnold.


Though Google declined Computerworld's request to discuss dataspaces or other ideas for which it has filed patent applications, Matthew Glotzbach, director of product management for Google's enterprise division, mentions the idea of profiling, the same tactic used to catch serial killers and other criminals. With search, empirical information would be used to create user profiles that could then be used to deliver "a really rich search experience."


Burton Group analyst Guy Creese opines that other search vendors will also begin to devote more effort to creating systems based on user profiles. So, for example, if a network engineer enters a query for "ATM," the system will rank results for "asynchronous transfer mode" more highly than "automated teller machine."

Add Comment      Leave a comment on this blog post

Post a comment





(Maximum characters: 1200). You have 1200 characters left.



Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.