Press Release

Attensity Workstation Debuts with Industry's Broadest Multi-Language Capabilities for Extracting and Transforming Unstructured Text for Integration with Structured Data

Palo Alto, Calif. - March 14, 2005 -- Attensity, the innovator in software applications that seamlessly integrate and analyze both structured and unstructured information, today announced Attensity Workstation, a powerful multi-language desktop addition to its family of industry-leading unstructured data transformation platforms. Attensity Workstation extracts and transforms facts captured in free-form text into structured data for easy integration with the tabular data of virtually any existing analytical software package. As the first Attensity product to incorporate the company's breakthrough new Directed LearningTM technology, Attensity Workstation offers the broadest multi-language capability of any product of its type, with the technology supporting virtually any single-byte language. In addition, Attensity Workstation represents a significant leap forward in ease of use, with a new example-based model that enables any user to master the skills for configuration with as little as 30 minutes of training, for "on the fly" analysis in response to time-critical requests.

Unstructured data - free-form text contained in emails, field reports, customer surveys, service records, warranty claims, Internet news feeds and other types of documents - is estimated to comprise 80% or more of most organizations' information. Attensity Workstation extracts relational facts and events that describe the "who, what, where, when and why" contained in text-based information, in multiple languages, and transforms this unstructured text into the structured, tabular data used by spreadsheets, relational databases and analytic applications. It enables organizations to generate analysis-ready datasets from free-form text sources rapidly, easily and accurately.

"Attensity Workstation is an important milestone on our product roadmap; our goal is to have multi-language capabilities in all our platform products," said Craig D. Norris, Attensity's chief executive officer. "With Attensity Workstation, we've made significant strides in terms of ease of use, enhanced language support and targeted text extraction capabilities. Never before has it been so easy for government agencies and other organizations to extract, transform, and integrate unstructured text-based information in multiple languages."

Attensity Workstation supports the most widely used operating systems. It integrates seamlessly with any business intelligence system or forms part of a larger solution from the Attensity Application Suite to support analytic applications, allowing users to extract from a broad set of data sources.

Intuitive, Example-Based Training
Unlike conventional text extraction products that may require configuration by a computational linguist or other specially trained expert, the new Directed LearningTM technology making its debut in Attensity Workstation enables ordinary users and analysts to quickly and easily "train" the system to extract and transform targeted facts and relationships into structured data for analysis.

After only a half-hour of instruction on Attensity Workstation, users can begin to easily and effectively "teach" the system how to find and extract the types of events needed from a corpus of documents. The user need only be conversant in the language of the document set and will then simply respond to examples derived by Attensity Workstation from the document set and displayed in its training interface. Based on the feedback received from the user, Attensity Workstation derives grammar and extraction rules, and is ready to be run over the data set to output structured data in a form that is easily loaded into the user's preferred repository, application or business intelligence solution. Attensity Workstation thus makes it simpler and faster for users to respond to ad hoc requests for time-sensitive analytical reports.

Language-Independent Capabilities
Attensity Workstation's Directed LearningTM technology is language-independent, capable of supporting text extraction in any language, with the broadest single-byte language support ever offered in any product of its type. Attensity Workstation supports most European (single-byte) and Middle Eastern (single-byte, bi-directional) languages. Asian languages (double-byte languages, such as Chinese, Japanese, and Korean), will be available in the second half of 2005.

According to Robert Lerner, senior analyst for application infrastructure at Virginia-based Current Analysis, "Attensity offers a powerful solution for unstructured data exploration by delivering software that's fast and flexible enough to make near-instant extractions from the seemingly endless amounts of data contained in both commercial and government environments. Perhaps the most notable aspects of the new Attensity Workstation, though, are its language-independence and user-friendliness - with Attensity Workstation, anyone in an organization can extract valuable information from unstructured data in any language within minutes of introduction to the product."

Product Availability
Available now, Attensity Workstation can be purchased directly from Attensity. Pricing information is available upon request; send inquiries to infows@attensity.com.

About Attensity
Attensity Corporation provides innovative applications that seamlessly integrate and analyze both structured and unstructured information for commercial enterprises and government agencies. Attensity's solutions are the first to fuse structured and unstructured data, enabling organizations to speed detection and response to critical events and issues that, until now, have been difficult to discover or resolve. Attensity's corporate headquarters are in Palo Alto, Calif., with its technology center in Salt Lake City, Utah. For more information, please visit www.attensity.com.