Sonic Foundry Announces Advanced Search Technology

Madison, WI — Sonic Foundry® Inc. (NASDAQ: SOFO), a leader in automated rich media communications technology, today announced significant advances in multi-modal search technology for the growing volumes of Mediasite™ libraries being generated by its customer base. The advancement now makes it faster and easier for organizations that are developing hundreds of hours of rich media presentations to efficiently find, access and playback specific information within that content.

Mediasite is a transformational communication medium for delivering critical information and sharing knowledge. The Mediasite webcasting and content management system automates the capture and delivery of presentations and lectures via the web.

Sonic Foundry’s new search capabilities combine advanced research and development in the areas of speech recognition, phonetic search, optical character recognition, language processing and contextual analysis into one all-encompassing search solution. Optical character recognition allows for quick word and phrase spotting within slide content or visual aids. Speech recognition, phonetic search, language processing and contextual analysis provide the ability to quickly and easily locate specific spoken words or phrases within a rich media archive, based on sophisticated algorithms developed by Sonic Foundry.

The company also has built a data processing capability that automatically extracts and creates contextually relevant metadata, eliminating the time and waste associated with manual metadata creation. Once Mediasite content has been indexed in multiple modalities, Sonic Foundry can then integrate the searchable index of information into custom search portals within customers’ enterprise networks. The result is an unprecedented method for automating the search of Mediasite content that provides tremendous additional benefits to Mediasite users.

Sonic Foundry has been steadily rolling out its advanced search capabilities as the market need accelerates for solutions that make it easy to identify and retrieve specific content within unstructured data files (streaming video, audio and accompanying graphical or image files). By controlling the capture and management of unstructured content, Sonic Foundry is able to optimize data structure and indexing for search algorithm efficiency and effectiveness. The company secured intellectual property pertaining to multi-modal search in late 2001 based on research originally conducted by Carnegie Mellon University in the early 1990s. The technology born out of that project was based on seven years of research and more than $20 million of funding by leading government agencies and private corporations, including NASA, DARPA, NSF, Bell Atlantic, Boeing, CNN, Intel and Microsoft. Sonic Foundry has continued to advance the technology to its current state based on additional research and development as well as customer input.

“We believe search lies at the heart of efficient, web-based communication. Finding a specific document or phrase has become a necessary part of working and learning,” explained Rimas Buinevicius, chairman and CEO of Sonic Foundry. “As our business grew and the amount of Mediasite content stored on our customers’ servers expanded exponentially, we knew users would face an increasing need to search that content for a particular piece of information. Organizations naturally want to leverage the communication they’ve captured, and this significant advancement, automatically creating searchable rich media databases for finding information instantly, further empowers them to reuse and repurpose that knowledge for the ongoing benefit of the organization.”