|
Courses and Workshops
|
| Title: | An introduction to spoken dialogue systems: fundamentals, standards and tools |
| Lecturer(s): | Ramón López-Cózar Delgado (Granada University) |
| Type: | Introductory Course |
| Section: | Language and Computation |
| Week: | First | | Time: | 11.00-12.30 (Slot 2) |
| Webpage: | http://www.ugr.es/~rlopezc
|
| Reader: | Download | | Room: | C1 |
Description Spoken dialogue systems are computer programs developed to provide
speech-based access to services such as train/airplane travel
information, weather information, automatic call routing, etc. This
type of system is currently employed by many companies and public
institutions to provide telephone-based information
automatically. Spoken dialogue systems are mainly based on three major
technologies. On the one hand speech processing, concerned with
recognising the words uttered by the user and generating
synthetic-voice responses. On the other hand linguistic processing,
concerned with obtaining the meaning of the recognised words and
generating syntactically and semantically correct text-based
responses. Finally, these systems are based on dialogue management,
concerned with enabling a natural interaction for users, solving
possible interaction errors and minimising the time required to obtain
the information or carry out the desired task. The course will
introduce these technologies to the students, analysing the
performance, problems and possible solutions for setting up the
diverse components of current systems. Finally, the course will
introduce several standards (VoiceXML, CCXML, etc.) and development
tools (CSLU Toolkit, Voxeo Designer, etc.).
PART 1. Introduction
1.1 Initial systems
1.2 Current systems
PART 2. Fundamentals of Spoken Dialogue Systems
2.1 Speech recognition
2.2 Natural language processing
2.3 Dialogue management
2.4 Natural language generation
2.5 Speech synthesis
PART 3. Standards
3.1 VoiceXML
3.2 CCXML
3.3 Other standards
PART 4. Development Tools
4.1 CSLU Toolkit
4.2 Voxeo Designer
4.3 Other tools
PREREQUISITES
Knowledge about high-level programming languages would be convenient
but it is not necessary.
lECTURER INFORMATION
Prof. Dr. Ramón López-Cózar Delgado is Associate Professor in the
Dept. of Languages and Computer Systems, Granada University (Spain).
In 1999 he read his Ph.D. Thesis entitled "Analysis, design,
implementation and evaluation of spoken dialogue systems. Development
of an experimental system".
He has published two books, one book chapter and more than 30 papers
in international and Spanish journals and conferences concerned with
spoken dialogue systems.
In 2001, he was lecturer of a the post-grade course "Man-machine
interaction using spoken dialogue systems", held at the National
Polytechnic University (Equator).
In 2003, he was lecturer in the post-grade course "Dialogue analysis
and management" held at the Dukes of Soria Foundation (Spain).
He is member of the COST Action 278 "Spoken Language Interaction in
Telecommunications" (European Commission).
He has participated in several research projects regarding dialogue
systems and voice transmission over IP.
His research interests focus on multimodal and multilingual spoken
dialogue systems, speech recognition and understanding, dialogue
management and speech synthesis.
|
© 2003, 2004 LORIA, Carlos Areces. |
2003-11-20
| |
