Cisco Blogs

Sweet Talk : Cisco and Speech Recognition

Note: This is the first of a two-part post.Today we conduct business around the clock and around the globe using seemingly infinite combinations of phones, voice messaging, e-mail, instant messaging, and video conferencing. As unified communications solutions integrate software, phones and computers, speech recognition promises to play an increasingly important role in the way we communicate, freeing our hands to let us control our experience with spoken commands instead of memorized, menu-driven clicks, keystrokes and button pressing. Yet, for all the promise of speech recognition, the technology has left much to be desired -until recently. One of the reasons that speech recognition solutions have failed in the past is that society has not developed a usage paradigm for a pure speech-based interaction with artificial intelligence. For example, when we call a speech-enabled customer service help line, we are not completely”trained” on the process, or paradigm. We are used to saying basic identifying information, such as an account number, and then prompted to say”balance,””transfer,” and similar commands because we do not know the specific words to say to this particular speech interface in order to check a balance. Sometimes we must navigate a system of menus and provide additional information to allow the system to better handle our query, but if things get complicated, we exit the speech session and talk to a real person to solve the problem. However, what happens when there is no person on the other end of the line acting as a safety net? Do we state our specific problem or do we provide some background information first? And if the system doesn’t understand us, do we repeat our statement or do we need to rephrase it? What are we supposed to say and how are we supposed to say it? This lack of a consistent and simple user interface is the second major reason that speech recognition solutions have failed to gain widespread traction.Cisco believes that in order for our customers to embrace and fully leverage speech recognition, the technology must offer both solution intelligence and a simple and natural user Mark Gervase, solutions marketing manager, Cisco Unified Communications

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.