Update: Recommind bolsters enterprise search

MIAMI – Recommind Inc. has upgraded its MindServer enterprise search software to make it easier for partners and system integrators to install and deploy, the company announced Monday.

MindServer, designed for corporations interested in categorizing, archiving and searching internal documents and retrieving information from them, now features in its 3.0 version an administration tool with a graphical interface that is intended to simplify the product’s configuration, said Recommind’s Chief Executive Officer Robert Tennant.

MindServer 3.0 also features APIs (application programming interfaces) and a more flexible and modular architecture that should make it easier to link the product with enterprise applications and information sources, he said. For example, MindServer 3.0 comes with prebuilt connectors for linking it to various portals, document and management systems, databases and text sources, he said. It also comes with prebuilt and customizable user interfaces for different applications, he said.

MindServer 3.0 also features increased scalability, to up to 100 million documents and thousands of users and enhanced security features, he said.

MindServer 3.0 is available now and starts at US$75,000, according to the Berkeley, California, company.

MindServer, which includes a retrieval component and a categorization component, is based on Probabilistic Latent Semantic Analysis algorithms, which allow the product to go beyond searches based on keywords and gain a conceptual understanding of documents, Tennant said.

MindServer supports over 150 document formats, including Microsoft Corp.’s Word, Excel and PowerPoint, Adobe Systems Inc.’s Acrobat and Corel Corp.’s WordPerfect, as well as e-mail attachments and documents formatted in HTML (Hypertext Markup Language).

“We try to provide access to all repositories of info in your organization and give you one interface to all those data sources.” he said.

The product targets vertical markets such as law, media and publishing, government and life sciences, and its clients include the international law firm Cleary, Gottlieb, Steen & Hamilton, global media giant Bertelsmann AG and the U.S. federal government’s National Library of Medicine.

The ability to search as broad a swath as possible of their internal documents is something companies increasingly want to do, across different systems, formats, repositories and platforms, said Sue Feldman, an IDC analyst. To meet this demand, systems integrators that previously focused on databases and transactional systems are broadening their expertise into enterprise search products, she said.

Thus, Recommind does well to make its product friendlier to systems integrators and other channel partners, she said. At the same time, simplifying the product’s use to novices inside the enterprise is also a good move for Recommind, she said. “This release appeals to the integrator as well as to the unsophisticated corporate user, and that’s very good,” she said.

The enterprise search market has big players such as Verity Inc., Autonomy Corp. plc, Fast Search & Transfer ASA and Convera Corp., but it doesn’t have any dominant players, she said. Moreover, it’s a market with solid annual growth, so smaller vendors with good technology and an astute approach have a chance to break away from the pack, she said.

Recommind has in its favor a solid product that focuses on concept-based searches and an emphasis on the growing legal segment, she said. “Recommind has as good a chance as any of becoming a major vendor,” she said.

“Demand for search among law firms is growing quickly because they need it for legal discovery, litigation and compliance. And they require a particular type of (very precise and broad) retrieval. They don’t need just an answer, or 80 percent of the answers, but all of the answers, and that’s difficult,” she said.

A product such as Recommind’s that is able to retrieve documents based on recognizing concepts, not just keywords, puts the vendor on the forefront of where search technology is moving, she said. “Language is ambiguous, and you can express the same idea in many different ways, so matching terms is hit or miss. The next wave is on identifying ideas, on searching for concepts, not on words,” she said.