ORION: from on-line interaction to off-line delegation

Seneff, Stephanie; Chuu, Chian; Cyphers, D. Scott

doi:10.21437/icslp.2000-229

Cited by 9 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The following dialogues contain several utterances in which context resolution must be performed to obtain a correct interpretation. Each dialogue takes place between a user and our spoken dialogue system in one of the GALAXY domains, which include JUPITER [53] for weather information, MERCURY [44] for flight reservations, ORION [41] for online task delegation, PEGASUS [54] for flight arrival and departure information, and VOYAGER [17] for Boston area traffic and landmark information.…”

Section: Relevant Phenomenamentioning

confidence: 99%

“…The intent of this thesis is to provide the reader with an understanding of the necessity for the new CR server and how it handles context resolution within our current domains. These include JUPITER [53] for weather information retrieval, MERCURY [44] for flight reservations, ORION [41] for online task delegation, PEGASUS [54] for flight arrival and departure information retrieval, and VOYAGER [17] for Boston area traffic and landmark information retrieval, as well as the simpler key-value-based domains created using the SPEECHBUILDER utility [19]. No meal is served on the first leg, and dinner is served on the second leg.…”

Section: Outlinementioning

confidence: 99%

“…In the GALAXY domains (JUPITER [53] for weather information, MERCURY [44] for flight reservations, ORION [41] for online task delegation, PEGASUS [54] for flight arrival and departure information, VOYAGER [17] for Boston area traffic and landmark information, and many others), the linguistic knowledge from an utterance is encoded in a hierarchical semantic frame containing objects identified as one of a small set of predefined types, namely, clause, topic, predicate, or keyword. TINA, the parsing component of the NL server, parses from a word graph produced by the speech recognizer and encodes the best hypothesis in a semantic frame representation.…”

Section: Galaxy's Knowledge Representationmentioning

confidence: 99%

See 2 more Smart Citations

A context resolution server for the galaxy conversational systems

Filisko¹,

Seneff²

2003

8th European Conference on Speech Communication and Technology (Eurospeech 2003)

Self Cite

View full text Add to dashboard Cite

The context resolution component of a spoken dialogue system is responsible for interpreting a user's utterance in the context of previously spoken user utterances, system initiatives, and world knowledge. This thesis describes a new and independent Context Resolution (CR) server for the GALAXY conversational system framework. The server handles all the functionality of the previous CR component in a more generic and powerful manner. Among this functionality is the inheritance and masking of historical information, as well as reference and ellipsis resolution. The new server, additionally, features a component which attempts to reconstruct the intention of a user in the case of a robust parse, in which some semantic concepts from an utter-I would also like to acknowledge the members of the Spoken Language Systems Group for their daily support. Specifically, I would like to thank Joe Polifroni for his substantial help in the initial stages of this project, and for his subsequent assistance in my understanding of discourse and dialogue. I also want to thank Scott Cyphers for his help whenever I had problems with the GALAXY code. Additional thanks go out to Grace Chung, Jim Glass, and Eugene Weinstein for their ideas and helpful testing of the server.I would like to thank my fellow graduate students for bearing with me the several times I attempted to elucidate the difference between discourse and dialogue.I want to give a special shout out to my office mates for making this an awesome and unforgettable year at MIT! Brooke, thanks for keeping us in line in our crowded, yet homey, office. Xiaolong, thanks for all your advice and for letting me teach you all those great idioms. And Vlad-sausage, pepperoni, and taekwondo. . . thanks, dawg.Last, but never least, I want to thank all my family and friends, whose words and deeds were inspirational and always appreciated. Manpreet K. Singh, MD, here's to the stick-to-itiveness of grad students! Andy, thanks for the ceol, damhsa, agus craic.Mom and Dad, thankyou always for your constant encouragement and unconditional support.

show abstract

Section: Relevant Phenomenamentioning

confidence: 99%

Section: Outlinementioning

confidence: 99%

Section: Galaxy's Knowledge Representationmentioning

confidence: 99%

See 1 more Smart Citation

A context resolution server for the galaxy conversational systems

Filisko¹,

Seneff²

2003

8th European Conference on Speech Communication and Technology (Eurospeech 2003)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Finally, there may be situations where the recognition performance could benefit from the addition of words not supported by the grammar. A case where this idea is well-motivated can be demonstrated by the ORION task delegation system [8]. ORION allows a user to leave a recorded message for later delivery.…”

Section: Additional Featuresmentioning

confidence: 99%

Automatic induction of n-gram language models from a natural language grammar

Seneff

Wang

Hazen

2003

8th European Conference on Speech Communication and Technology (Eurospeech 2003)

Self Cite

View full text Add to dashboard Cite

This paper details our work in developing a technique which can automatically generate class n-gram language models from natural language (NL) grammars in dialogue systems. The procedure eliminates the need for double maintenance of the recognizer language model and NL grammar. The resulting language model adopts the standard class n-gram framework for computational efficiency. Moreover, both the n-gram classes and training sentences are enhanced with semantic/syntactic tags defined in the NL grammar, such that the trained language model preserves the distinctive statistics associated with different word senses. We have applied this approach in several different domains and languages, and have evaluated it on our most mature dialogue systems to assess its competitiveness with preexisting n-gram language models. The speech recognition performances with the new language model are in fact the best we have achieved in both the JUPITER weather domain and the MERCURY flight reservation domain.

show abstract

The use of subword linguistic modeling for multiple tasks in speech recognition

Seneff

2004

Speech Communication

Self Cite

View full text Add to dashboard Cite

ORION: from on-line interaction to off-line delegation

Cited by 9 publications

References 4 publications

A context resolution server for the galaxy conversational systems

A context resolution server for the galaxy conversational systems

Automatic induction of n-gram language models from a natural language grammar

The use of subword linguistic modeling for multiple tasks in speech recognition

Contact Info

Product

Resources

About