Internet Communications

Can speech-recognition software break down educational language barriers?
Answered by Vinton G. Cerf, Rob Wrubel and 1 other
  • Vinton G. Cerf

    Vinton G. Cerf

  • Rob Wrubel

    Rob Wrubel

  • Susan Sherwood

    Susan Sherwood

  1. Vinton G. Cerf Vice President and Chief Internet Evangelist, Google


    [A new] trend is the ability for computers to… Let me use the word "recognize speech," rather than saying "understand it" because that's a big [artificial intelligence] problem. But recognizing spoken speech and being able to turn it into text: Once you can do that, you can use that for command and control or for query and getting answers back.

    If you're illiterate, [it's also] a really big help because you can now talk to the Internet and talk to the Web. But even more important, once you get it in text form, you can use translation techniques to go from one language to another. Google recently announced the ability to actually do that. Imagine sitting with your laptop or your mobile [phone], speaking in English, and having the person on the other end hear German. Well, I won't argue that the German is perfect, and there won't be misunderstandings as a consequence. We have misunderstandings with our friends in England because we define the same word differently. But for the most part, we're seeing these technologies beginning to overcome barriers to communication.


    More answers from Vinton G. Cerf »

  2. Rob Wrubel Chief Marketing & Product Development Officer, Apollo Group


    Well, in the end, when people learn together, it's the most powerful place to actually connect societies and cultures. And when you do it in the process of learning, whether it's thin film deposition techniques and IT guys in China, India and the U.S. are all obsessed -- their ideological differences, all sorts of differences dissolve in the process of that exchange. I am a big believer that educational networks that span the globe and connect us by our interests and our desires to add value to the world ultimately are a social good that dissolves the ideological issues, objections, or religious objections -- the things that hold us back that sometimes are being packaged and framed in such a way that they probably are artificial.

    More answers from Rob Wrubel »

  3. Speech-recognition software has been used in many areas. Let's look at its educational impact on foreign language learning. For several years, language learning software has been available with an automatic speech recognition (ASR) feature. Using it mirrors the practice of traditional language labs, but feedback comes from a computer rather than an instructor. The software is being used in colleges and universities for on-line courses, traditional classes that include web-based content and entire language labs. Consumers can also purchase the programs for home use.

    The voice recognition software component is one aspect of comprehensive, computer-based language learning. ASR is used during a variety of exercises: Student word and speech pronunciation is compared to that of a native speaker. The students receive continual assessment and are informed of the accuracy of their language production. Assignments are repeated numerous times and progress is monitored.

    In 2004, the professional journal "Language Learning and Technology" published a favorable review of the ASR element of specific language learning software. Within this program, student pronunciation is assessed by several metrics, including amplitude and pitch. Feedback is offered in formats such as graphs, scores and computerized replies. The reviewers considered the voice recognition component extremely accurate.

    The journal article did note some time limitations with the ASR. In order for the software to accurately evaluate language, students cannot answer until they hear a beep signal. Following that, they have only five seconds to complete their response.

    In general, language learners respond favorably to voice recognition software. A 2009 study from the University of South Florida reported that students using ASR during foreign language study believed the voice recognition software accurately assessed their language production, and they found the subsequent feedback valuable. In fact, a majority of students found ASR analysis more beneficial than other methods of critique. Learners thought that, compared to instructor monitoring, ASR offered more opportunities for repetition and more extensive evaluation of pronunciation.

    For many students, learning a foreign language is a difficult process. Speech-recognition software is a new addition to the language toolbox that is improving students' pronunciation and winning them over as well.

    More answers from Susan Sherwood »

Still Curious?
  • What are the benefits of Yelp membership?

    Answered by HowStuffWorks

  • What is a local area network?

    Answered by Science Channel

  • What is Google Maps?

    Answered by HowStuffWorks


What are you curious about?

Image Gallery