MUE2007 :: 2007 International Conference on Multimedia and Ubiquitous Engineering ::

Tutorial

2007 International Conference on Multimedia and Ubiquitous Engineering

April 26th ~ 28th, 2007
Korean Bible University , Seoul, Korea

1. Dr. Byeongho Kang (University of Tasmania, Australia)
2. Dr. Dominik Slezak (Infobright Inc., Canada)
3. Dr. Daniel Howard (QinetiQ, UK)

	'Push' in Web Information Management: Limitations and Possibilities
	Dr. Byeongho Kang Professor, University of Tasmania, Australia
Summary of His Talk
The major problem of the current WWW technology is that it is developed based on the 'pull' style of information delivery, where the uploaded information waits for visitors, and not the 'push' style, where the new information is delivered to the users when it becomes available. In future WWW services and ubiquitous environments, the fast delivery of newly available information and re-use of archived information will be primary characteristics of information delivery systems. The 'push' delivery was introduced in early internet service and it is widely used by internet users. However, the main problems in 'push' are information overflow (spam) by information providers and the lack of personalization for each user. In this talk, I would like to present 'push' based information services with the automated classification method. A prototype system has been used as a state government in Australia. The system has monitored about 300 Australian government Web pages for operators of the official government Web pages. The classification system has been tested by about 20 users and the result shows promising performance in the acquisition of classification knowledge from human operators. The main characteristic of the classification system is an incremental learning technique so called as Multiple Classification Ripple Down Rules. The system should be adaptable to changes in the classification of documents or introduction of new classifications. Also, we will address mobile delivery systems for collected information.
About Dr. Byeongho Kang
Dr. Byeong Ho Knag is a senior lecturer at the School of Computing, University of Tasmania, Australia and a head of Research and Development of a joint venture company, kmAgent Pty.Ltd.,Hobart Australia. He received his Ph.D from the University of New South Wales, Sydney in 1996 and has worked in the Advanced Research Lab. HITACHI, Japan and Hoseo Univerity, Korea, before he joined the University of Tasmania in 2000. He also has worked in research and development projects with industries and research organizations, the Smart Internet Collaborative Research Centre, in Australia, the Asian Office of Aerospace Research Department, US in Japan. His research includes basic Knowledge Acquisition methods and many applied researches in Internet systems as well as medical expert systems. His work appears to be one of the few examples of research that has gone beyond criticizing the dominant assumptions and attempted to develop and apply different types of systems based on alternate assumptions. He has been involved in the development of several commercial AI products, an expert system development tool, an Intelligent Help Desk System, Web based information monitoring and classification systems, etc. He has been involved in the foundation of several joint venture companies in Australia and Korea. At present, the Tasmanian State Library is using his Web monitoring system for the Tasmanian government Web site. He has published many papers in refereed journals, conference proceedings and book chapters in these areas and has been involved in many conference and workshops as conference chair or program committee member. He recently organized the 19th Australian Joint Conference on Artificial Intelligence 2006 in Hobart, the Pacific Knowledge Acquisition Workshop in China 2006 and in New Zealand 2004.