• Create BookmarkCreate Bookmark
  • Create Note or TagCreate Note or Tag
  • PrintPrint
Share this Page URL
Help

Part: II The Process of Designing Speech-Recognition Systems

Part II: The Process of Designing Speech-Recognition Systems

In Part I we explored the business case for automating data access and transactions over the telephone. And we covered the functional advantages of speech-recognition over touchtone technology. I also introduced some high-level design issues that concern us when we make a speech-recognition application.

As you'll see, there's a whole lot more to the process than simply figuring out which number on the keypad to assign to a given response. A speech-recognition system requires research—to learn as much as possible about the client's business and its customers. It requires a comprehensive, coherent design that incorporates every facet of the insights gained through the research, as well as an understanding of the psychology of how people interact with computers. Designs are embodied in Design Specifications, in which the most significant aspect are the prompts that must be written effectively if they are to create a successful application. The design and development must capture the client's branding requirements. The application requires rigorous usability testing—both before and after deployment—to ensure that it has been properly designed and implemented so that real callers can (and will want to) use it. Underlying all of these issues is the design philosophy. We can't achieve any of the other aspects of good design without a strong philosophical basis. We'll discuss these aspects from the designer's perspective in the chapters to come—as it's often the designer who assumes the responsibilities for these many roles.



PREVIEW

                                                                          

Not a subscriber?

Start A Free Trial


  
  • Creative Edge
  • Create BookmarkCreate Bookmark
  • Create Note or TagCreate Note or Tag
  • PrintPrint