Chapter 4. Multimodal Speech Applications

The World Wide Web exploded in the late 1990s, culminating in the presence of Web sites for nearly every company. Since then Web sites have been continually stabilized and enhanced. Web-based applications are often the preferred method of access to information. It is not enough to offer a dynamic Web site. To attract new users, you must offer something unique and customer oriented.

Chapter 3 utilized the Microsoft Speech SDK (SASDK) to create a telephony, or voice-only, application. The SASDK can also be used to develop what are known as multimodal applications. These are similar to traditional Web-based applications, but they are unique in that they offer the user a choice of input mechanisms.

Multimodal applications execute on an IIS Web server and are rendered by clients running the Speech add-in for Microsoft's Internet Explorer. There is also an add-in available for PocketPC Internet Explorer, so you can easily create speech-enabled mobile applications. What makes these applications unique is that the user is in control and can choose to utilize speech by clicking a button or icon. At that point, speech text can be used to navigate the application more efficiently than traditional point-and-click methods.

This chapter will examine a multimodal application designed for a fictional community college. The college wishes to offer students an alternative and high-tech way of registering for the upcoming semester.

The chapter also features a case study involving a real company named Landstar systems. Landstar has used the SASDK to build a multimodal application that it expects will increase call-routing efficiency and reduce agent turnover.