Self-Organizing Dialogue Management

In this paper, we present our approach to dialogue management in the spoken dialogue system that is being developed within the project Interact. Compared to traditional approaches, our dialogue manager will support the system’s adaptivity and flexibility with the help of two design decisions: an agent-based architecture and the use of neural network models. Our experiments focus on word-based dialogue act recognition using the LVQ classification algorithm in a corpus of information-seeking dialogues, and we compare the results with a simple bag-of-words approach. We also report our studies of clustering the input data into necessary and meaningful categories using self-organizing maps.

[1]  Norbert Reithinger,et al.  Utilizing Statistical Dialogue Act Processing in Verbrnobil , 1995, ACL.

[2]  Arne Jönsson,et al.  A model for habitable and efficient dialogue management for natural language interaction , 1997, Natural Language Engineering.

[3]  Stefan Wermter,et al.  An Overview of Hybrid Neural Systems , 1998, Hybrid Neural Systems.

[4]  Stefan Wermter,et al.  SCREEN: learning a flat syntactic and semantic spoken language analysis using artificial neural networks , 1997 .

[5]  Markku Turunen,et al.  Jaspis - a framework for multilingual adaptive speech applications , 2000, INTERSPEECH.

[6]  JönssonArne.  A model for habitable and efficient dialogue management for natural language interaction , 1997 .

[7]  Timo Honkela,et al.  WEBSOM - Self-organizing maps of document collections , 1998, Neurocomputing.

[8]  Kristiina Jokinen,et al.  Confidence-Based Adaptivity in Response Generation for a Spoken Dialogue System , 2001, SIGDIAL Workshop.

[9]  Stefan Wermter,et al.  A Novel Modular Neural Architecture for Rule-Based and Similarity-Based Reasoning , 1998, Hybrid Neural Systems.

[10]  James F. Allen,et al.  An architecture for a generic dialogue shell , 2000, Natural Language Engineering.

[11]  Kimmo Koskenniemi,et al.  A General Computational Model for Word-Form Recognition and Production , 1984 .

[12]  Jorma Laaksonen,et al.  LVQ_PAK: The Learning Vector Quantization Program Package , 1996 .

[13]  Risto Miikkulainen,et al.  Subsymbolic natural language processing - an integrated model of scripts, lexicon, and memory , 1993, Neural network modeling and connectionism.

[14]  Norbert Reithinger,et al.  Modeling Negotiation Dialogs , 2000 .

[15]  Philippe Bretier,et al.  ARTIMIS: Natural Dialogue Meets Rational Agency , 1997, IJCAI.

[16]  Timo Honkela,et al.  Self-Organizing Maps In Natural Language Processing , 1997 .

[17]  Kristiina Jokinen,et al.  Goal Formulation based on Communicative Principles , 1996, COLING.

[18]  David R. Traum,et al.  Cooperation, dialogue and ethics , 2000, Int. J. Hum. Comput. Stud..

[19]  Marilyn A. Walker,et al.  Automatic Optimization of Dialogue Management , 2000, COLING.

[20]  Jens-Uwe Möller.  Using DIA-MOLE for unsupervised learning of domain-specific dialogue acts from spontaneous language , 1991 .

[21]  Kristiina Jokinen.  Learning Dialogue Systems , 2000 .

[22]  Timo Honkela,et al.  Self-Organizing Maps in Symbol Processing , 1998, Hybrid Neural Systems.

[23]  Stefan Wermter,et al.  Towards Hybrid Neural Learning Internet Agents , 1998, Hybrid Neural Systems.