Teragram Entities and Events Extraction Technology Keeps America Up-To-Date by Delivering Customized Breaking News

Text mining technology enables major news outlets to provide immediate, personalized news content as it happens

CAMBRIDGE, Mass. - June 20, 2005 - Teragram, the leading provider of multilingual natural language processing technologies, announced today that its entities and events extraction platform powers the leading subscription- based news alerts offered by the nation's top media outlets.

As readers of major U.S. news portals know, the ability to tailor news content for their specific needs and interests is more important today than ever before. Due to the sheer volume of international news sources now available, keeping up-to-date on breaking issues, especially those that readers must be aware of for their jobs, academic needs or personal interests, can be challenging. For this reason, the use of customized alerts continues to grow dramatically.

"In a recent study conducted by Delphi we found that 42 percent of business people spend a day or more each week looking for information needed to do their job," said Carl Frappaolo, industry analyst at Delphi Group, a Perot Systems Company. "The use of content classification tools is becoming a must-have in virtually every industry. The volume of content in portals, intranets and internets mandates organization via a taxonomy scheme."

Teragram's text mining technology, which includes entities and events extraction, helps people and companies solve this problem and is behind the breaking news alerts of many media outlets, including New York Times Digital and CNN.com. Registered users of these sites can specify topics, key words and public company ticker symbols to be monitored for the instant delivery of breaking news. As soon as news occurs that matches customized criteria, it is immediately pushed to the subscriber's desktop as a breaking e-mail alert. This convenient tool provides users with a personalized news delivery service while also helping to bring them back to the news outlet's Web site, a critical business function.

As news arrives at New York Times Digital and CNN.com it is electronically routed through Teragram's software, where the technology extracts concepts and key events and categorizes the information into taxonomies. It is at this stage that Teragram's entities and events extraction engine stands out above typical key word search tools as it can categorize by concepts and related subject matter for more accurate and relevant output. When a content match occurs based on a subscriber's defined criteria, Teragram triggers an alert, keeping the reader informed. Additionally, recipients can specify not only the news they want to receive, but how and when they wish to receive their alerts. Teragram's technology even offers suggested additional articles and links based on the user's alert choices.

Teragram's Market-Leading Entities and Events Extractor Capabilities

The Teragram Entities and Events Extractor automatically detects and extracts entities, concepts and events from text and documents. Entities and concepts such as people's and company's names, publicly traded businesses, titles and positions, and geographical locations are among many of the concepts that are automatically extracted from text. These concepts can be easily customized for specific domains, organizations and uses. This product is available as an OEM product to enhance other applications such as CRM, Knowledge Base, and Search, or as a stand-alone enterprise solution.

In addition to identifying concepts in text, Teragram Entities and Events Extractor can also output information associated with the extracted concepts. For example, Teragram Concept Extractor not only extracts publicly traded companies mentioned in documents but also outputs their ticker symbol and the stock market where they are listed.

The Extractor utilizes Teragram's Automatic XML Marker technology to enable the automatic production of XML tags associated with concepts extracted by the Extractor technology.

About Teragram
Teragram Corporation is the market leader in multilingual natural language processing technologies that use the meaning of text to distill relevant information from vast amounts of data. Founded in 1997 by innovators in the field of computational linguistics, Teragram alone offers the speed, accuracy and global language support that customers and partners demand to retrieve and organize growing volumes of digital information. Teragram helps customers perform more efficient searches and better organize information in more than 30 languages, enabling them to reach new markets and make better decisions. Teragram serves customers across the publishing, pharmaceutical, telecommunications and financial industries, including major news organizations, leading online search portals, and many Fortune 1000 companies. Customers include among others Ariba, Ask Jeeves, CNN, Factiva, FAST Search & Transfer, Forbes.com, InfoSpace, Kofax, NYTimes Digital, OneSource, Ricoh, Sony, Verity, WashingtonPost.com, the World Bank, and Yahoo. For more information please contact 1-617-576-6800 or visit http://www.teragram.com/info