Teresa Sikorski

Learn More
This paper describes a system that leads us to believe in the feasibility of constructing natural spoken dialogue systems in task-oriented domains. It specifically addresses the issue of robust interpretation of speech in the presence of recognition errors. Robustness is achieved by a combination of statistical error post-correction, syntactically-and(More)
This paper describes a system that leads us to believe in the feasibility of constructing natural spoken dialogue systems in task-oriented domains. It specifically addresses the issue of robust interpretation of speech in the presence of recognition errors. Robustness is achieved by a combination of statistical error post-correction, syntactically-and(More)
In this paper we describe a recent experiment designed to evaluate the performance of the TRAINS-95 system. The evaluation uses a task-based evaluation methodology appropriate for dialogue systems such as TRAINS-95, where a human and a computer interact and collaborate to solve a given problem. In task-based evaluations, techniques are measured in terms of(More)
Data collection and annotation represent the most significant portion of the effort involved in building statistical models of dialogue. We must be able to annotate reliably if there is any hope of building predictive statistical models. This paper describes the annotation process being used to tag the TRAINS-93 corpus with dialogue acts, and measures we(More)
  • 1