In Search of...
How does the newest enterprise search technology work?
Endeca Guided Navigation dynamically helps users sift through giant data sets—whether they’re wine cellars or data warehouses.
Guided Navigation instantly analyzes the thousands of search results to generate relevant categories that can help you narrow your search—instead of simply overwhelming you with 2,917 hits.
Guided Navigation generates only meaningful next steps for refining your search, never a dead end. For example, since all search results at this stage are wines from only two countries, only United States and South Africa are offered as Country categories.
Guided Navigation creates sensible and relevant value ranges for quantitative parameters like Price Range—so searchers quickly understand what’s available to them, and avoid fruitless searches for unavailable results.
Because it always pairs search results with relevant categories for refinement, Guided Navigation pulls more value out of data—not only helping users find things faster, but also revealing other ways to think about the data (like Ratings, Drinkability, Flavors) and choices they may not know about.
The Endeca revolution is powered by a paradigm shift in technology. Plain vanilla Web sites use relational databases, search engines, or rigid navigation systems to let visitors navigate offerings. All three tools have problems. They overwhelm users with unwieldy lists of results, or return the frustrating “no results found,” without indicating where to go from there.
For instance, if you type “history” in the search box at Amazon.com’s bookstore section, the query chokes the user with 32,000 results. “That’s because it doesn’t know if ‘history’ is part of a title or a description,” explains Endeca’s product marketing manager, Peter Bell. When users try to dig further they often come up empty “because there are so few possible ways to access each record. So the catalog is essentially invisible.”
Endeca technology, on the other hand, creates hundreds of browse paths to each record in a given data set. And it doesn’t keep all that juicy information to itself; instead it organizes and displays it dynamically, without losing sight of the original query. Web sites powered by InFront and ProFind, Endeca’s two major search and navigation products, solve the opposing problems of information overload and queries ending abruptly with zero results.
To experience this improvement, browse Barnes & Noble’s Web site for the Endeca version of searching for “history.” The results are stunning. Three subsets immediately appear: nonfiction, fiction and children. Each subset is organized by a host of subtopics, from the obvious (European) to more obscure (cooking, parenting and science fiction). Click under “children” and new subtopics appear: the child’s age range, featured authors, and a bunch of others, like Black U.S. history—and it’s all for kids. This is a whole new ball game: users can find books through this interface that they might never have known existed.
How does Endeca work its magic? “Our navigation engine starts with all the records in a given data set, whether that’s a catalog or a list of mutual funds,” says Bell, “and works backward to build out every valid path to each record.”
It’s like creating a book with a wonderfully detailed and accurate index. The fact that the Endeca software builds the index ahead of time makes it extremely fast; results pop up immediately. That all dead ends are eliminated makes for a far more satisfying search-and-discover user experience—whether shopping online or analyzing internal company email@example.com
Maintained by: firstname.lastname@example.org
Last modified: Aug 31, 2004, 17:07 EDT