| Paper: | SLP-P6.8 | ||
| Session: | Speech Understanding, Translation, Applications and Systems | ||
| Time: | Tuesday, May 16, 16:30 - 18:30 | ||
| Presentation: | Poster | ||
| Topic: | Speech and Spoken Language Processing: Spoken Language Applications and Systems | ||
| Title: | WEBTALK: TOWARDS AUTOMATICALLY BUILDING SPOKEN DIALOG SYSTEMS THROUGH MINING WEBSITES | ||
| Authors: | Junlan Feng, Dilek Hakkani-Tür, Giuseppe Di Fabbrizio, Mazin Gilbert, Mark Beutnagel, AT&T Labs – Research, United States | ||
| Abstract: | WebTalk is a system for analyzing unstructured information from company websites to support automatic creation of spoken dialog applications. The goal is to completely automate the process of building, maintaining and deploying these applications by leveraging the wealth of information on given websites. WebTalk attempts to do that by exploiting technologies in web mining, document understanding, question/answering, and speech and language processing. In this paper, we review extensions to those technologies that can make them suitable for creating a WebTalk application without the need for human intervention. We present an evaluation study of a WebTalk spoken dialog system that has been instantiated on a telecom company website. Experiments with 30 different scenarios indicate promising results and provide evidence that such systems can potentially revolutionize the paradigm of creating and scaling spoken dialog services. | ||
©2018 Conference Management Services, Inc. -||- email: webmaster@icassp2006.org -||- Last updated Friday, August 17, 2012