Conference Paper

Let's go public! taking a spoken dialog system to the real world.

Conference: INTERSPEECH 2005 - Eurospeech, 9th European Conference on Speech Communication and Technology, Lisbon, Portugal, September 4-8, 2005
Source: DBLP

ABSTRACT In this paper, we describe how a research spoken dialog system was made available to the general public. The Let's Go Public spoken dialog system provides bus schedule information to the Pittsburgh population during off-peak times. This paper describes the changes necessary to make the system usable for the general public and presents analysis of the calls and strategies we have used to ensure high performance.

1 Follower
 · 
77 Views
  • Source
    • "Our experiments are based on the Let's Go corpus (Raux et al., 2005). Let's Go contains recorded interactions between a spoken dialogue system and human users who make enquiries about the bus schedule in Pittsburgh. "
    SIGDIAL; 01/2013
  • Source
    • "The context was the Noctívago system, an experimental agenda-based SDS in European Portuguese that provides schedule information about night buses in Lisbon. The first version of the system was adapted from Let's Go [7], a live system that gives bus schedule information for real users in Pittsburgh since 2005. Both telephone systems were based on the Olympus open-source architecture for SDSs [8], and used Ravenclaw [9], an agenda-based dialog manager. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper proposes an approach to the use of lexical entrainment in Spoken Dialog Systems. This approach aims to increase the dialog success rate by adapting the lexical choices of the system to the user's lexical choices. If the system finds that the users lexical choice degrades the performance, it will try to establish a new conceptual pact, proposing other words that the user may adopt, in order to be more successful in task completion. The approach was implemented and tested in two different systems. Tests showed a relative dialog estimated error rate reduction of 10% and a relative reduction in the average number of turns per session of 6%.
    Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on; 01/2013
  • Source
    • "These tools are used worldwide and are known to work on a variety of languages, dialects and tasks. They are also used for rapid creation of speech recognition and synthesis systems for new languages [2], in resource-scarce conditions [3], in memory crunched mobile devices [4], and in deployed real world applications [5]. The Festvox environment has also been used for teaching graduate-level Speech Technology [6] "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper summarises the tools provided within Festvox[1], a freely available software suite for creation and analyses of large scale speech corpora for enabling research, development and instruction in speech technologies.
Show more

Preview

Download
1 Download
Available from