News

Big Blue open-sources speech code

• By John K. Waters
• September 21, 2004

IBM is donating some of its software for speech-enabling applications to two open-source organizations: the Apache Software Foundation and the Eclipse Foundation. Big Blue announced its plans at last week's SpeechTEK 2004 Conference to "spur the availability of speech-enabled applications by making it easier and more attractive for developers to build and add speech recognition capability in a standardized way."

To the Apache group, IBM is donating Reusable Dialog Components (RDCs), which are JSP tags developed by IBM Research. The RDCs enable dynamic development of voice applications and multimodal user interfaces. "Multimodal" refers to applications that provide users with a choice of input sources, generally including voice, keypad, keyboard, mouse and stylus. Output takes the form of spoken prompts, audio and/or graphical displays. JSPs that incorporate RDC tags automatically generate W3C VoiceXML 2.0 at runtime, providing a standard basis for speech applications.

To Eclipse, IBM is donating speech mark-up editors designed to make it easier for developers to write standards-based speech applications, and to create and utilize RDCs within those applications. The Eclipse donation is actually part of the initial formation of a project for open-source tools for voice application development, which will involve several other companies in the VoiceXML community, the company says.

The two donations combined are valued at $10 million, an IBM spokesperson says.

Some of the companies joining IBM in this open-source initiative include AT&T, Avaya, Cisco, Genesys, Motorola, Nortel, ScanSoft, Siebel and Voice Partners.

IBM has been an active, longtime supporter of open-source development. The company participates in and contributes to more than 150 open-source projects, says Jon Prial, VP of marketing in IBM's information management group. Those contributions include a $1 billion investment in Linux in 1998 and a $40 million donation to Eclipse in 2001 (IBM actually launched Eclipse). Just last month, the company donated its Cloudscape Java-based, embedded database to the Apache group. At the time, Prial said of the company's open-source strategy: "It's all about building the broader developer community. [With an open-source project] you get innovation, you get a broader community working on technology, and it allows us to innovate at the technology level and above it."

IBM is also no stranger to the speech-recognition business. The company introduced its first software-only speech recognition product, Voicetype, in 1996. It later created Viavoice, the first "continuous speech" engine. About five years ago, the company released the Viavoice Toolkit for Linux.

The donations were widely seen as a strategic marshalling of open-source community resources to speed development of IBM-friendly applications that compete with Microsoft. In March, Microsoft released Microsoft Speech Server 2004 for speech-enabled applications. According to the firm, more than 100,000 programmers have downloaded Microsoft's free developers' kit for building speech applications on the .NET platform.