IBM WebSphere Voice Server Brings Transcription to the Enterprise

Large Vocabulary, Server-Based Dictation Technology is Easy for Developers to Customize

Select a topic or year

Somers, NY - 29 Jan 2002: IBM today announced new speech recognition technology for companies to implement advanced, enterprise-wide dictation and transcription. Based on IBM's award-winning ViaVoice technology, it allows developers to build and customize solutions for specific industries. Professionals, such as doctors and lawyers, can then dictate medical reports or legal records into a device or telephone and have their work returned as text - reducing the reliance on labor-and cost-intensive human transcription. Enrollment is required only once, after which users can dictate from virtually any location within the network.

The new software, IBM WebSphere Voice Server for Transcription, is already being incorporated into industry-specific solutions. With a vocabulary of more than 160,000 words, it is designed for multiple users - whether in a small law firm or a large corporation. Also available are additional, specialized vocabularies for medical and legal fields.

Because it is part of IBM's extensive WebSphere family of infrastructure software, developers can link the transcription server to the rest of the enterprise. The Topic Factory, a built-in tool, also allows developers to add specialty vocabularies quickly and easily. Network administrators also can tune voice profiles for increased accuracy. Dictation is done via devices including telephones, USB microphones and mobile digital recorders, and then processed on a server.

Traditional transcription is labor-intensive and costly. Voice files are recorded and sent to live agents, who painstakingly transcribe, edit, and send the text back. Transcription using speech technology, on the other hand, generates text with greater accuracy and speed, leaving only the editing to the agent. Also, professional transcriptionists, especially those with specialized industry knowledge, are hard to find. Speech recognition helps transcription providers make more efficient use of their resources.

Early adopters of IBM's Transcription Server include Austin, Texas-based medical solutions provider, MD Productivity, today announced MD One - one of the first commercially available telephone-based medical transcription solutions using speech technology. MD One allows physicians to dictate their records or reports over practically any device including a simple telephone. It is then sent over the server and transcribed electronically, after which it is edited by a professional transcriptionist and sent back to the physicians via e-mail.

"With MD One, physicians do not have to learn anything new or change their habits. They simply dictate as they have for the past 20 years - our technology takes care of everything else," said Newt Hamlin, Chief Executive Officer of MD Productivity. "By incorporating voice technology, we are able to make more efficient use of a limited resource - transcriptionists - while enabling physicians to work more productively."

Another developer, Berlin-based DictaNet, has built a German and English language solution aimed primarily at the law profession. Lawyers dictate into a digital mobile recorder or a PC and send audio files to DictaNet where it is transcribed using speech recognition. The text can then be edited by either the lawyer, or a transcriptionist. DictaNet plans to expand its service beyond the legal profession.

"Speech technology has steadily made its way into business applications especially over the past two years, giving companies valuable tools to increase efficiency," said William S "Ozzie" Osborne, General Manager, IBM Voice Systems. "The transcription server allows professionals to use their most natural form of communication - speech - to work more productively."

The Transcription Server is packaged with a runtime server, a toolkit, APIs, programming and system administration documents, utilities, and sample client code.

Related XML feeds
Topics XML feeds
Information Management, Lotus, Tivoli, Rational, WebSphere, Open standards, open source