COLING 2004

Satellite Workshop

 

 

 

 

Robust and Adaptive Information Processing

for

Mobile Speech Interfaces

 

DUMAS Final Workshop

 

 

 

Edited by

Björn Gambäck and Kristiina Jokinen

 

 

 

August 28-29, 2004


Geneva Switzerland



Robust and adaptive information processing for mobile speech interfaces

Introduction to the workshop

 

Kristiina Jokinen

Björn Gambäck

University of Helsinki and

University of Art and Design Helsinki

Hämeentie 135 C

FIN – 00560 Helsinki, Finland

kjokinen@uiah.fi

SICS

Swedish Institute of Computer Science AB

Box 1263

SE – 164 29 Kista, Sweden

gamback@sics.se


 


The workshop focuses on adaptation and mobile technology in intelligent interactive systems. Especially, advanced research on machine learning techniques and robust technology needed in building adaptive systems will be discussed and demonstrated. The workshop is organized around central two themes:

·         Robust speech and language processing in practical applications

·         Future challenges for adaptive interaction technology

 

The state of the art speech and language technology is already on such a high level that the users can execute simple control commands to direct system operations, and also have short conversations with the system to search for information. However, to enable natural multilingual interaction, the unique requirements that arise from the combination of adaptive interface design, dialogue research, and language processing have to be addressed in the system development.

To this end, we encouraged submissions investigating adaptive multilingual interaction techniques to handle both spoken and text input and to provide coordinated linguistic responses to the user. In particular papers concerning innovative frameworks and offering new, even controversial view-points for developing interactive systems. The overall aim of the workshop is to provide a forum for discussing and showing how adaptive information processing can be combined in spoken dialogue systems to offer natural and rich interaction with various users; the second day of the workshop ends with a panel discussion on the future challenges in the field, while the first day ends with live demonstrations of working systems.

A predecessor of the workshop was co-organized at the European Chapter for ACL in Budapest last year. The present workshop focuses on the scientific results and opportunities provided by the current state of the art in speech and language technology. It also functions as the final workshop of the EU-IST project DUMAS, and provides an opportunity for combining high-quality research (presented both in the workshop and the preceding main conference) with concrete results and practical development prospects for human language technology.

The workshop is meant to cover both academic researchers as well as industrial partners. The target audience consists of researchers and system developers who share an interest in intelligent spoken dialogue systems and want to know about recent advances in robust speech and language processing. It is expected that the workshop will also have relevance to industrial partners, and draw their interest to advances in speech and language technology research in general. In particular, three invited speakers together provide connections to other IST projects, to the industry, and to the end users of spoken language systems.

We look forward to fruitful discussions and interesting demonstrations, and wish you welcome to the final workshop of the DUMAS project.



Workshop Chairs and Programme Committee

The workshop is co-chaired by Björn Gambäck, SICS (Administrative Coordinator of DUMAS) and Kristiina Jokinen, University of Art and Design Helsinki (Scientific Coordinator of DUMAS).

 

The scientific committee for the workshop consists of the following experts:

 

Alex Bailey, Canon Research Centre Europe,UK

Bill Black, UMIST, England

Peter Boda, Nokia Research Center, Finland

Rolf Carlson, KTH, Sweden

Roberta Catizone, University of Sheffield, UK

Robert Dale, Macquarie University, Australia

Laila Dybkjaer, University of Southern Denmark, Denmark

Michael McTear, University of Ulster, Ireland

Victor Poznanski, Sharp Laboratories of Europe, UK

Allan Ramsay, UMIST, UK

Frank Reh, DaimlerChrysler, Germany

Uli Strempel, Etex, Germany

Alistair Sutcliffe, UMIST, UK

Pasi Tapanainen, Connexor, Finland

Markku Turunen University of Tampere, Finland

Tom Wachtel, Independent Consultant, Italy/UK

Yorick Wilks, University of Sheffield, UK

 

 

 

Workshop website: http://mlab.uiah.fi/~kjokinen/coling-se1.html

 

COLING website: http://www.issco.unige.ch/coling2004/index.html

 

 

DUMAS website: http://www.sics.se/dumas/


Workshop Schedule

Saturday August 28

09:15 Welcome

09:30 – 10:30 Invited Speaker: Michael McTear (University of Ulster, Newtownabbey):

New Directions in Spoken Dialogue Technology for Pervasive Interfaces

10:30 – 11:00 Markku Turunen et al. (DUMAS project):

AthosMail: A Multilingual Adaptive Spoken Dialogue System for an E-mail Domain

 

11:00 – 11:30 Coffee

 

11:30 – 12:00 Gambäck, Olsson, Sahlgren, Funk, Ramsay, Paajanen (DUMAS project):

                        The AthosMail Text Processor

12:00 – 12:30 Black, Conroy, Funk, Ramsay, Stairmand and Thompson (UMIST, Manchester):

Multilingual Discourse Processing

 

12:30 – 13:00 Discussion

 

13:00 – 14:00 Lunch

 

14:00 – 15:00 Invited Speaker: Peter Brass (German Federation of Blind and Visually Impaired People)

15:00 – 15:30 Jokinen, Kanto,  Kerminen and Rissanen (University of Art and Design Helsinki):

Evaluation of Adaptivity and User Expertise in a Speech-based E-mail System

 

15:30 – 16:00 Coffee

 

16:00 – 18:00 DUMAS demonstrations and posters:

·         Multilingual AthosMail (Turunen et al.)

·         AthosNews for Finnish (Heiskari)

·         AthosNews for English (Stairmand)

·         The DUMAS Calendar Application (Pakucs)

·         Robust Language Analysis (Järvinen et al.)

·         Speech Applications at ETeX (Schöbel-Lunshof)

18:00 Reception

 


Sunday August 29

09:30 – 10:00 Invited Speaker: David Horowitz (Vox Generation, London):

                    FASiL: Flexible and Adaptive Spoken Language and Multi-modal Interface

10:00 – 10:30 Panttaja, Reitter and Cummins (Media Lab Europe; University College Dublin):

The Evaluation of Adaptable Multimodal System Outputs

10:30 – 11:00 Alex Bailey (Canon Research, London):

Challenges and Opportunities for Interaction on Mobile Devices

 

11:00 – 11:30 Coffee

 

11:30 – 12:00 Invited Speaker: Péter Pál Boda (Nokia Research, Helsinki):

Modality Integration in a Wider Sense

12:00 – 12:30 Cheng, Calvedon and Dale (CSLI, Stanford; Macquarie University, Sydney):

Generating Navigation Information Based on the Driver’s Route Knowledge

12:30 – 13:00 Antti Kerminen and Magnus Sahlgren (UIAH, Helsinki; SICS, Stockholm):

Vector-based DACT Classification

 

13:00 – 14:00 Lunch

 

14:00 – 14:15 Tom Wachtel (UMIST, Manchester):

Video: “Do You Do Dumas?”

14:15 – 15:45 Panel discussion: Future Challenges

Alex Bailey, William Black, Peter Boda, Robert Dale, Björn Gambäck,

David Horowitz, Kristiina Jokinen, Michael McTear, Victor Poznanski,

Helmut Schöbel-Lunshof, Yorick Wilks

 

15:45 – 16:00 End of the workshop and coffee

 

DUMAS Meeting


Table of Contents

 

Introduction to the Workshop.......................................................................................................... 3

Workshop Chairs and Programme Commitee................................................................................... 4

Workshop Schedule....................................................................................................................... 5

Table of Contents.......................................................................................................................... 7

Acknowledgements........................................................................................................................ 8

 

 

Alex Bailey

Challenges and Opportunities for Interaction on Mobile Devices............................................... 9

 

William Black, Andrew Conroy, Adam Funk, Allan Ramsay, Mark Stairmand, Paul Thompson

Multilingual Discourse Processing............................................................................................. 15

 

Péter Pál Boda

Modality Integration in a Wider Sense....................................................................................... 22

 

Hua Cheng, Lawrence Calvedon, Robert Dale

Generating Navigation Information Based on the Driver’s Route Knowledge............................ 31

 

Björn Gambäck, Fredrik Olsson, Magnus Sahlgren, Adam Funk, Allan Ramsay, Sirkku Paajanen

The AthosMail Text Processor.................................................................................................... 39

 

Petteri Heiskari

The Finnish AthosNews: a News-reading Application for Visually Impaired Users.................... 43

 

Kristiina Jokinen, Kari Kanto,  Antti Kerminen, Jyrki Rissanen

Evaluation of Adaptivity and User Expertise in a Speech-based E-mail System.......................... 44

 

Timo Järvinen, Mikko Laari, Timo Lahtinen, Sirkku Paajanen, Pirkko Paljakka,

Mirkka Soininen, Pasi Tapanainen

Robust Language Analysis Components for Practical Applications............................................ 53

 

Michael McTear

New Directions in Spoken Dialogue Technology for Pervasive Interfaces................................. 57

 

Botond Pakucs, Sanna Huhta

Developing Speech Interfaces for Frequent Users: The DUMAS Calendar Prototype............... 65

 

Erin Marie Panttaja, David Reitter, Fred Cummins

The Evaluation of Adaptable Multimodal System Outputs.......................................................... 69

 

Markku Turunen, E-P. Salonen, M. Hartikainen, J. Hakulinen, W. Black, A. Ramsay,

A. Funk, A. Conroy, P. Thompson, M. Stairmand, K. Jokinen, J. Rissanen, K. Kanto,

A. Kerminen, B. Gambäck, M. Sahlgren, F. Olsson, M. Cheadle, P. Hansen, S. Nylander

AthosMail: a Multilingual Adaptive Spoken Dialogue System for an E-mail Domain................. 77

 

 

Author Index............................................................................................................................... 87

 


Acknowledgements

 

The workshop is sponsored by the EU IST project DUMAS (Dynamic Universal Mobility for Adaptive Speech Interfaces), IST-2000-29452. Thanks to all the project partners for discussions and collaboration on the work, and to the European Commission for financial support.

 

We would like to thank our programme committee for reviewing the submissions and helping us to make the decisions for the final programme. We also thank the COLING Workshop Chairs Michael Hess and Fabio Rinaldi, as well as the local conference organisers for their help in practical organisation.

 

 

Björn Gambäck

 

Kristiina Jokinen