An online prototype speech-enabled information access tool using Java Speech Application Programming Interface

The World Wide Web today has become a major force in driving the economy, and the technology. In spite of the abundance of information freely available on the Web, the information is not accessible to everyone. There are a couple of reasons for this trend. Firstly, most of this information has primarily been used by and restricted to persons who connect and use the Web via a traditional computer. Thus access to Web information, for an increasing number of mobile device users, is very limited at this point. Secondly, most of the contents on the Web are in a format unsuitable for disabled people. Contents are either visually rich or highly structured. Proposed here is the prototype of an application tool that would provide increased access to Web information using speech technology. The implemented tool is based on distributed computing techniques to provide scalability. The tool incorporates both speech recognition and speech synthesis thereby providing complete speech input and output capability, tending towards a complete hands-free Web access. The proposed tool is implemented using a Java Speech Application Programming Interface (JSAPI) compliant speech engine which uses Java as the host language. This arrangement provides platform independence and portability to the tool. The tool is based on client-server architecture and uses remote method invocation (RMI) for distributed computational purposes. This paper provides a brief overview of the design and implementation of the application, includes discussion on current research issues and provides observations and suggestions for future enhancements.