Developing a Flexible Spoken Dialog System Using Simulation

Developing A Flexible Spoken Dialog System Using Simulation Grace Chung � Corporation for National Research Initiatives 1895 Preston White Drive, Suite 100 Reston, VA, 20191 gchung@cnri.reston.va.us Abstract In this paper, we describe a new methodology to develop mixed-initiative spoken dialog systems, which is based on the extensive use of simulations to accelerate the development process. With the help of simulations, a system providing informa- tion about a database of nearly 1000 restaurants in the Boston area has been developed. The simula- tor can produce thousands of unique dialogs which benefit not only dialog development but also pro- vide data to train the speech recognizer and under- standing components, in preparation for real user interactions. Also described is a strategy for creat- ing cooperative responses to user queries, incorpo- rating an intelligent language generation capability that produces content-dependent verbal descriptions of listed items. 1 Introduction Spoken dialog systems have traditionally been dif- ficult to instantiate because of the extensive efforts required for configuring each component from the natural language (NL) understanding to the domain- specific context tracking and dialog engines. This task is particularly involved when building systems that empower users with greater flexibility at the spoken input via mixed-initiative interaction (Zue et al., 2000; Ferguson and Allen, 1998), and systems that enable querying across large and frequently changing databases. The goal of this work is to assemble natural spo- ken dialog interfaces that enable flexible interac- tions through mixed initiative dialog and coopera- tive responses. Such interfaces aim to help users navigate large information spaces such as on-line databases. Conventional systems generally guide the users through a series of scripted prompts, either through � This research is sponsored in part by SPAWAR SSC-SD. The content of this paper does not necessarily reflect the position or policy of the Government, and no official endorsement should be inferred. User: Are there any Thai restaurants? System: Can you provide a neighborhood or city? User: Boston. System: There are 21 entries found. Can you be more specific? User: A restaurant in Dorchester. System: There are no Thai restaurants in Dorch- ester. Is there anything else I can help you with? User: What about in the South End? System: There are 2 entries found. Thai Village and House of Siam. Table 1: Example human-computer dialog. a finite-state grammar or a dialog control table. In Table 1, the system prompts present a sequence of questions in an attempt to solicit more constraints from the user, with the goal of obtaining a small data subset from the database. The system responses are generated from a set of rules that only anticipate one of a handful of situations: (1) when the set of en- tries returned is too large, (2) the set of entries is adequately small to enumerate, and (3) no available entries have been returned. A more flexible scenario would allow the user to browse the content by specifying one or more con- straints in any order. The system should then re- turn a succinct summary …

Content truncated to 500 words