Recent Improvements in the CMU Spoken Language Understanding System

RECENT IMPROVEMENTS IN THE CMU SPOKEN LANGUAGE UNDERSTANDING SYSTEM Wayne Ward and Sunil Issar School of Computer Science Carnegie Mellon University Pittsburgh PA 15217 ABSTRACT We have been developing a spoken language system to recognize and understand spontaneous speech. It is difficult for such systems to achieve good coverage of the lexicon and grammar that subjects might use because spontaneous speech often contains disfluencies and ungrammatical constructions. Our goal is to respond appropri- ately to input, even though coverage is not complete. The natural language component of our system is oriented toward the extraction of information relevant to a task, and seeks to directly optimize the correctness of the extracted information (and therefore the system response). We use a flexible frame-based parser, which parses as much of the input as possible. This approach leads both to high accuracy and robustness. We have implemented a version of this system for the Air Travel Information Service (ATIS) task, which is being used by several ARPA-funded sites to develop and evaluate speech understanding systems. Users are asked to perform a task that requires getting information from an Air Travel database. In this paper, we describe recent improvements in our system resulting from our efforts to improve the coverage given a limited amount of training data. These improvements address a number of problems including generating an adequate lexicon and grammar for the rec- ognizer, generating and generalizing an appropriate grammar for the parser, and dealing with ambiguous parses. 1. INTRODUCTION Understanding spontaneous speech presents several problems that are not found either in recognizing read speech or in parsing written text. Since the users are not familiar with the lexicon and grammar used by the system, it is very difficult for a speech understanding sys- tem to achieve good coverage of the lexicon and grammar that sub- jects might use. Spontaneous speech often contains ungrammatical constructions, stutters, filled pauses, restarts, repeats, interjections, etc. This causes problems both for the recognizer and the parser. Stochastic language models tend to produce more robust recognition than grammar based models. They can be smoothed to allow for unseen word sequences and their scope is short enough to "get back on track" after an error. The parsing and understanding component also must be robust to the phenomena in spontaneous speech and to recognition errors. Even though the speech is disfluent and gramat- ically ill-formed, the relevant information is still consistent most of the time. We therefore try to model the information in an utterance rather than its grammatical structure. The natural language compo- nent of our system is oriented toward the extraction of information relevant to a task, and seeks to directly optimize the correctness of the extracted information (and therefore the system response). We use a flexible frame-based parser, which parses as much of the input as possible. This approach leads both to high accuracy and robustness. We have implemented a version of this system for the ARPA Air Travel Information Service (ATIS) task. Users are asked to perform …

Content truncated to 500 words