BHP stores documents on current and past projects in the central repository of its content-management system, but it found that its metadata-based search engine didn't discover all significant documents related to a particular project or line of inquiry (in part because authors hadn't been consistent in applying metadata). Nor did the engine take into account structured data available in well reports and geologic databases.
"The problem is that in certain cases, a lot of the information may lie in a database, but in other cases, it may just be a graph in a PowerPoint presentation," says Michael Glinsky, section leader in quantitative interpretation at BHP and the leader of the company's recent search project.
Early this year, integrator Blue Fish Development Group helped BHP deploy Endeca ProFind search tool on BHP's intranet. Rather than relying on existing taxonomy or data structure, which can limit possible paths to records related to a particular search, ProFind explores every valid pathway to results and then presents "guided search" contextual and classification information alongside the conventional hit list. The system can break out separate pathways just to the articles, photographs, tables, audio, or maps related to a search. This lets BHP engineers quickly refine searches and drill down into smaller, more relevant collections, rather than browsing through long lists of unrefined results, Glinsky says. The approach also helps engineers working on projects discover similar projects and proven solutions.
BHP's system isn't passive. When projects are launched and related documents are checked into the content-management system, Endeca automatically searches for projects with similar characteristics and pushes that information to engineers. BHP spent about $1 million on the project, an amount the company expects to recover quickly because the system will enable engineers to identify more accurately oil fields with the greatest potential for successful production.
"We had a mountain of information and no effective way to go through it," Glinsky says. "Now we're getting better value out of millions of dollars of research and influencing decisions that can create hundreds of millions of dollars in revenue."
Design-automation software firm Mentor Graphics Corp. gained broad user acceptance of its enterprisewide search implementation by targeting a pain point. The company used to rely on separate search functions on the intranet and the consumer Web site, depending on whether staff or customers wanted to search the knowledge base, online documentation, or internal document repositories. These siloed search tools made it tough to find answers to technical questions. When Mentor purchased a competitor and needed to upgrade its customer-support infrastructure early last year, it looked for a single search tool that would support internal needs and Web customer self-service.
After narrowing the field to four products, Mentor tested each on its own content by posing a set of 1,500 questions and evaluating the answers. After reviewing the results, the company chose InQuira to power its SupportNet online system. InQuira's software handles content and structured data and includes a semantic processing engine that's said to understand and respond to intent even if the user is struggling to find the proper search terms.
Mentor's internal users were frustrated with the separate search silos, so they readily accepted the new system. Customer response also has been positive, says Tom Floodeen, VP of worldwide customer support, pointing to increased Web traffic as evidence. Mentor will spend about $6 million less on call-center services this year thanks to the SupportNet upgrade, Floodeen says, yet the total project cost $4 million.
Ultimately, the goal of search is simple: to find answers and make better decisions. Putting people in touch with more sources of information, including content and data as well as internal and external sources, is a step toward information awareness. To get to information intelligence--the ability to find the right information and associations that might not be obvious at first glance--users are experimenting with text mining, data extraction, clustering, guided navigation, and other techniques that are helping make overload a thing of the past.