Seeking far and wide for the right data

Offering enterprises hope for digesting the exploding volume of internal and external content, search-and-categorization technologies are pushing into enterprises to become essential elements of corporate portal infrastructure to power other content-driven applications.

First brought to prominence by the rise of the Web, search technology is now emerging as a valuable tool for calming the enterprise data storm, according to analyst Guy Creese, research director at Aberdeen Group Inc., in Boston.

“Search first became important because of the Web. The Web made people understand what search was,” Creese says. “Search was unsexy about two years ago … but in the past year search technologies have just blossomed.”

A solid search and categorization infrastructure that spans enterprise portals and other applications is a vital weapon for handling the deluge of the content-pounding corporation, Creese says.

“[Search] is an important issue for enterprises because there is so much content now. If [companies] don’t have a sophisticated way of finding that content, they are blowing their investments in the content,” Creese says. “Search is becoming much more of an infrastructure play, [highlighting] how it works with other applications, like Web analytics or CRM. It is a tool you can apply in many different ways.”

Illustrating the significance of organizing and finding information within enterprise portals, search software maker AltaVista and infrastructure software vendors Autonomy and Verity this month agreed to provide search-and-automation functionality for IBM Corp.’s WebSphere Portal Server.

In a similar vein, AltaVista recently announced a deal with Oracle intended to inject search capabilities into Oracle’s enterprise-portal initiative. Under the deal, AltaVista created two portal applications for Oracle 9iAS, designed to allow Oracle portal users the ability to access both Web-based search and customized enterprise-search functionality.

According to Gary Bryan, director of business development at Autonomy Corp PLC, based in Cambridge, England, the capability of Autonomy’s infrastructure software to comprehend unstructured data is a valuable contribution to portal and other application environments.

“Once you have an understanding of the data itself, you can perform operations on it such as search,” Bryan says. “To us, search is just one of the things necessary for a total portal environment. Most portal environments give you a window into different data stores — such as a Notes database, an e-mail repository, or an Oracle database — but don’t tie them together. [Autonomy] can get an understanding at the data level of what is in the data, and automatically relate those things together.”

Meanwhile, search technology vendor iPhrase is developing connectors designed to bridge its natural language search-and-navigation platform with corporate portal offerings from vendors such as Plumtree and Epicentric Inc., according to Tony Frazier, vice president of product marketing at iPhrase.

The Cambridge, Mass-based company’s One Step platform combines natural language understanding, information retrieval, and user-interaction algorithms to interpret search requests and query all relevant data sources, officials said.

In addition to integration partnerships, several search-based infrastructure products have recently entered the market.

Sunnyvale, Calif.-based Verity Inc. last month rolled out K2E (K2 Enterprise), infrastructure software designed to integrate search and categorization into corporate portals and intranets.

K2E’s Social Networks feature helps people interact with data and documents they create, access, and search, according to Rajat Mukherjee, principal software architect at Verity.

“In corporations, it is not just data interacting with data, but people interacting with data,” Mukherjee says. The ability to use business rules is critical to making taxonomies and data more appropriate and useful.”

Another vendor making headway with search functionality is Santa Clara, Calif.-based Smartlogik Group PLC, which recently unveiled two software products designed to help enterprises find and use their content.

Smartlogik’s Muscatstructure is a rules-based categorization engine that enables the development of taxonomies personalized to individual users, according to Smartlogik officials. Smartlogik’s natural-language search tool, Muscatdiscovery, lets users ask a question to locate content within a corporation or partner network, or from external sources, such as the Internet.

The company’s software is designed to run on Internet-based application server platforms, including Sun’s Java J2EE (Java 2 Enterprise Edition) and Microsoft’s Com+, which can aid the process of linking the software to other corporate applications and systems, according to Smartlogik officials.

The combination of search and categorization is a powerful tool for enterprises, says Campbell Macpherson, marketing director at Smartlogik.

“It depends on the industry, but access to the right info at the right time lets you bring a drug to market sooner in the pharmaceutical industry, for example,” Macpherson says.

“No one buys search for [its own] sake; it is a business case to be put together on an individual basis,” Macpherson says. “It can increase productivity considerably by finding the right info quickly and [ultimately] improve ROI.”