Home

General Information
Registration
Program Schedule
Accommodation
Call for Papers
Call for Workshops
Call for Tutorials
Call for Doctoral Forum
Accepted Papers & Keynotes
Conference Committee
About New Zealand

Sponsored By

 





 
 

2004 PRICAI Pre-Conference
Tutorials

The following tutorial will be held during the conference of PRICAI 2004. You are cordially invited to visit these well-designed homepages and explore the possibility for you to actively participate in one of the tutorials interesting to you.

All tutorial are held on 9th & 10th August. The morning and afternoon sessions are roughly 9:00am - 1:00pm and 2:00pm - 5:30pm. All the tutorials will be held at the PRICAI 2004 conference site. Please click on these tutorials for their important dates and other details.

Tutorial One: Language Technology: Applications and Techniques
Tutorial Two: Introduction to speech processing
Tutorial Three: Relevant Reasoning: Its key role in discovery and prediction
Tutorial Four: Specifying knowledge and belief change in multiagent systems
 


 
Tutorial 1 Language Technology: Applications and Techniques
     
Presenter: Prof Robert Dale
Date: 9th August 2004 Morning Session
     

Context and Motivation
Language Technology is the new millennium's practically-focussed rebirth of natural language processing, covering applications from optical character recognition to sophisticated spoken language dialog systems and intelligent search engines. Language Technology is widely perceived by the IT industry to be a fundamental enabling technology that will both enable smarter interfaces and provide assistance in overcoming the information overload of the Internet age.

Tutorial Aims and Content
This tutorial aims to provide the attendee with a broad awareness of actual and potential Language Technology applications, along with a framework for thinking about these applications in terms of the linguistic resources they need. Attendees will acquire an understanding of the scale of development required for different kinds of applications, along with an appreciation of what constitutes a feasible application. Frequent reference will be made to commercial applications, with corresponding critiques aimed at showing how to assess claims made by vendors.

The tutorial will cover the following topics:

Applications of Language Technology:
Spoken Language Dialog Systems
Machine Translation
Text Summarisation
Search and Information Retrieval
Question-answering Systems

Techniques for Natural Language Processing:
Word level information
Syntax and semantics
Statistical processing

Biography of Presenter
Robert Dale has an international research reputation in natural language processing, and particularly in natural language generation; he has presented numerous tutorials on these topics at international conferences.

He is author of over 60 journal and conference papers, and is author and editor of a number of books in the area; most recently Building Natural Language Generation Systems (Reiter and Dale 2000; Cambridge University Press) and the Handbook of Natural Language Processing: Tools and Techniques (Dale, Moisl and Somers [eds] 2000, Dekker Publishing). He is the Director of the Centre for Language at Macquarie University, and editor of the journal Computational Linguistics.

He is also Director of Language Technology Pty Ltd, a consultancy focussing on cutting-edge speech and language applications. He writes a regular column entitled Industry Watch in the Journal of Natural Language Engineering.

Tutorial 2 Introduction to speech processing
     
Presenters: Dr Catherine Watson
  Dr Waleed Abulla  
Date: 9th August 2004 Afternoon Session
     
Context and Motivation
Speech Recognition, and Speech Synthesis are now main stream technologies, and Spoken Language systems are on the verge of being so. However there are still many hurdles to cross in achieving natural spoken language systems. This is due not only to computational limitations but also because of the complexities of speech itself, both linguistically and in terms of the signal. One of the fascinating aspects of speech processing is that knowledge of the linguistic features, perceptual models and signal processing techniques are all essential for successful processing.

Tutorial Aims and Content:
This tutorial aims
to provide a broad awareness of acoustic features of speech and how speech can be parameterised and of actual and potential applications of speech processing. It is to provide an understanding of the importance of speech corpora, along with an application case study: building a speech recogniser. Attendees will acquire an understanding of the complexities involved in speech processing, along with an appreciation of how this technology is applied. There will be a particular focus on speech recognition.

The tutorial will cover the following topics

Speech Processing
• acoustic phonetic description of speech
• perception of speech
• parameterisation of the speech signal

Applications of Speech Processing (topics include)
• speech recognition
• speech synthesis
• speech training

Building Speech Corpora
• Data Collection
• Data Storage
• Data Obsolescence

Speech Processing Case Study: Building a Speech Recognizer
• Feature extraction
• pattern matching
• training
• testing

Brief Bio of the Presenters
Catherine Watson is an Engineer by training, and has recently joined the department of Electrical and Computing Engineering at the University of Auckland. However she has also spent 5 years in the Linguistics department at Macquarie University, Sydney, Australia, as a member of the Speech, Hearing, Language Research Centre, and 3 years in a joint position between the department of Electronics and the Macquarie Centre of Cognitive Science, also at Macquarie University. This has provide her with the unique opportunity to study speech from many angles. She is the author of over 30 journal and conference papers, including one in Nature about the Queen’s English.

Dr. Abdulla has a PhD degree from the University of Otago, Dunedin, New Zealand. He was awarded the Otago University Scholarship for 3 years and the bridging grant. He is currently working as a Senior Lecturer in the department of Electric al and Computer Engineering/The University of Auckland. His main research area is in Digital Signal Processing for
non-stationary signals, Statistical Modelling, and Artificial Neural Networks. He is a visiting researcher to Siena University/Italy. He has collaborative work with Essex University/UK. He has more than 35 publications and currently supervising 9 postgraduate students.



Tutorial 3 Relevant Reasoning: Its key role in discovery and prediction
     
Presenter: Jingde Cheng
Date: 10th August 2004 Morning Session
     

Context and Motivation
Many research problems in Artificial Intelligence are concerned primarily with drawing new conclusions from given premises rather than finding a justification for an explicitly specified statement from given premises. The most intrinsic difference between reasoning and proving is that the former is intrinsically prescriptive and predictive while the latter is intrinsically descriptive and non-predictive. The purpose of reasoning is to find some new conclusion previously unknown or unrecognized, while the purpose of proving is to find a justification for some specified statement previously given. Proving has an explicitly given target as its goal while reasoning does not. Unfortunately, until now, many studies in Computer Science and Artificial Intelligence disciplines still confuse proving with reasoning. Discovery is the process to find out or bring to light of that which was previously unknown. Prediction is the action to make some future event known in advance, especially on the basis of special knowledge, and therefore, it is a notion must relate to a point of time to be considered as the reference time.

For any discovery and/or prediction, both the discovered and/or predicted thing and its truth must be unknown before the completion of discovery and/or prediction process. Since reasoning is the only way to draw new, previously unknown conclusions from given premises, there is no discovery and/or prediction process that does not invoke reasoning. Moreover, to be effective and efficient, any reasoning in discovery and prediction should be relevant such that there must be some connection of meaning, i.e. some relevance, between its premises and its conclusion.

Tutorial Aims and Content
This tutorial provides an informal and elementary introduction to relevant reasoning and relevant logic for computer scientists from the viewpoint of their applications in discovery and prediction. Its purpose is to show that truth-preserving and relevant reasoning based on relevant logic can play many important roles in Knowledge Science as well as Computer Science, and therefore, to call attentions of computer scientists to research directions and challenge problems on relevant reasoning and its applications. After giving philosophical definitions of reasoning, proving, discovery, prediction, logic, and the notion of conditional, we give a historical review of the background and motivation of relevant logic, survey the state of the
art of relevant logic in both proof theory and model theory, show applications of relevant reasoning based on strong relevant logic in Knowledge Science as well as Computer Science, and point out some research directions and challenge problems.

Biography of presenter
Jingde Cheng is a professor of computer science at Saitama University in Japan. His current research interests include ampliative reasoning and relevant reasoning, relevant logic and its applications, epistemic programming for scientific discovery, autonomous evolution of knowledge-based systems, anticipatory systems, and formal methods in information security engineering. He has published over 160 refereed papers in journals, books, and conference proceedings. He received the degree of Bachelor of Engineering in computer science from Tsinghua University in China in 1982, and the degree of Master of Engineering in computer science and the degree of Doctor of Engineering in computer science from Kyushu University in Japan in 1986 and 1989, respectively. He is a member of ACM, IEEE, IEEE-CS, IEEE-SMS, AAAI, AAR, IPSJ, and JSSST.


Tutorial 4 Specifying knowledge and belief change in multiagent systems
     
Presenters: Hans van Ditmarsch
Date: 10th August 2004 Afternoon Session
     
Context and Motivation
Architectures of multiagent systems can be seen as a specific kind of interpreted or distributive system, where interacting processors or agents at least have knowledge of their internal state, but where the knowledge or beliefs they have of other agents, varies. Over the past twenty years also many results have been achieved to model the dynamics of such systems. This generally involves introducing additional operators describing the effect of actions. These operators can be seen as 'epistemic programs'. This has led to advances in knowledge-based protocol verification, analysis of game states and game actions involving knowledge, and belief revision for model-based agents.

Tutorial Aims and Content:
The tutorial is intended to give people working in multiagent systems some additional formal tools for their specification. The target audience are graduate students active in the area, and researchers active in related areas. Familiarity with propositional logic is a requirement. Familiarity with modal logic is not a requirement but an advantage. Extensive examples and in-class training are part of the tutorial.

In the setting of multiagent system architectures the concepts of 'knowledge' and 'belief' have a precise technical meaning, both on a structural and on a descriptive level. Knowledge is assumed to satisfy specific properties, such as that everything that you know is true, that you are aware of your knowledge, and that you are also aware of your ignorance. Epistemic logic is the modal logic of knowledge, where the modality means 'knowing that'. We will present this logic, also for more than one agent, and also for additional modal operators such as 'common
knowledge'. For examples we tend to use concrete multiagent systems such as card games.

The dynamics of such systems will be presented in detail. We focus on ways to specify the results of public announcements. Game actions - that are seen as game state transformers - and various communicative interactions can be made precise in the dynamic extensions of the logics presented. Puzzling phenomena such as statements that become false when announced, have a clear explanation in this setting. The logic is also suitable for the analysis of knowledge-based security protocols, where sending and receiving agents attempt to keep secrets from eavesdroppers.

'Belief' is generally taken to mean something slightly weaker than knowledge. We will give examples of 'belief' in the context of card games, and of the interaction of knowledge and belief.

The logics presented have sometimes peculiar theoretical properties, such as axiomatizations that require rather advanced semantic proof techniques to establish their completeness, and untractable complexities for basic proof search tasks. Such issues of fascinating interest for an audience of theoretical computer scientists will not be addressed in this tutorial but will only be mentioned in passing.

Brief Bio of the Presenters
Hans van Ditmarsch completed an MSc in Mathematics and in Philosophy at the University of Utrecht in 1986. He then lectured and worked in course development at various Dutch universities. In 1996 he started and in 2000 he completed a PhD in Mathematics and Computer Science at the University of Groningen, under the supervision of Johan van Benthem (Amsterdam and Stanford) and Gerard Renardel. The topic of his PhD is a language for dynamic epistemic logic and applications of that language for modelling games. He joined the Computer Science Department of the University of Otago in 2001. His current research focusses on the dynamics of knowledge, combinatorics, and computer and information science education. He has written various textbooks and has been actively publishing in his research area in the years since the completion of his PhD. He has given graduate courses at ESSLLI 2003 in Vienna, at the Summer School in Logic and Automated Reasoning 2003 in Canberra, and at the (upcoming) EASSS 2004 in Liverpool.

   
 
     
   
© 2003-2004 PRICAI. All rights reserved. Site designed by Jaytech
     
Home Sitemap