Intelligent Media Agents: Key Technology for Interactive Television, Multimedia and Internet Applications

ISBN-13: 9783528057060 / Angielski / Miękka / 1999 / 180 str.

Hartmut Wittig

Opis

Intelligent agents are rescuer in the information glut. They help users to find information which better corresponds to their interests and needs. This book describes the architecture and basic modules of an intelligent media agent. A personal television guide is described as an example of intelligent help, addressing the problem of managing TV channels by using an intelligent agent.
Contents: intelligent agents - software agents - programming - fuzzy logic
Target groups: Computer scientists - students - software developers - multimedia and internet programmer
About the Author: Hartmut Wittig is directing the corporate strategy department of "Deutsche Telekom Computer Service Management" (DeTeCSM). In the recent years, H.Wittig has directed multimedia and internet software R&D departments and groups. He gave lectures on "Multimedia Systems" in German universities and worked as Editor in Chief of IEEE Multimedia Newsletter.

Spis treści

1 Introduction.- 1.1 Research in Interactive Television Systems.- 1.2 Applications.- 1.2.1 Television and Interactive Television.- 1.2.2 Value-Added Services.- 1.3 The Problem of an Information Glut.- 1.3.1 Motivation for TV Program Selection.- 1.3.2 Program Search in Television Systems.- 1.4 Need for Agent-based Systems.- 1.5 Overview.- 2 Towards Personal Television Systems.- 2.1 Software Agent Definition.- 2.1.1 Agency.- 2.1.2 Intelligence.- 2.1.3 Mobility.- 2.2 Related Work.- 2.3 Practical Issues and Examples.- 2.3.1 Adoption of User Interfaces/Dialogs.- 2.3.2 Shopping and Other Interactive Services Support.- 2.3.3 Software Agents for Information Retrieval.- 2.3.4 Personal EPG and Movie Generation.- 2.4 Intelligent Agent Research Applied to Television.- 2.4.1 Movie Select.- 2.4.2 Movie Recommendation System (MORSE).- 2.4.3 Siemens Personal Newspaper.- 2.4.4 Evaluation.- 3 Requirements.- 3.1 Requirements to Multimedia Environments.- 3.1.1 Customizing.- 3.1.2 Portability.- 3.1.3 Modifiability.- 3.1.4 Multiple Application Support.- 3.1.5 Distributed vs. Local Processing.- 3.2 Requirements to Program Provision.- 3.2.1 Program Providers.- 3.2.2 Program Description.- 3.2.3 Fuzzy Specification.- 3.3 Requirements to User-Specific Functions.- 3.3.1 Specification of User Demands.- 3.3.2 Processing and Output.- 3.3.3 Privacy and Security Issues.- 3.4 Conclusion.- 4 Agent Run-Time Environment.- 4.1 System Standards.- 4.1.1 DAVIC.- 4.1.2 DVB.- 4.1.3 Multimedia Standards in DAVIC and DVB Compared.- 4.2 Multimedia Standards.- 4.2.1 Multimedia and Hypermedia Experts Group Standard.- 4.2.2 HTML and JAVA.- 4.3 Dynamic Generation of Multimedia Applications.- 4.3.1 Virtual Object Store.- 4.3.2 Location of VOS.- 4.3.3 Comparison.- 5 Architecture of the Intelligent Media Agent.- 5.1 Multilayered Agent Approach.- 5.2 Module Description.- 5.2.1 Multimedia User Interface.- 5.2.2 Basic Agent Operating System.- 5.2.3 High-level Softbots.- 5.2.4 Communication Environment.- 5.2.5 Architecture.- 5.3 Discussion of Privacy and Security Aspects.- 6 Program Descriptors and User Profiles.- 6.1 Classification Units.- 6.2 Program Descriptors.- 6.2.1 Statistics.- 6.2.2 Genre.- 6.3 User Profile.- 6.3.1 Identification and Privacy Aspects.- 6.3.2 Desires and Demands.- 6.3.3 Time of Day, Day of Week, Holiday.- 6.3.4 Fuzzy Attributes.- 6.4 Attribute Types.- 7 Matching and Clustering.- 7.1 Matching Methods.- 7.1.1 Defmition.- 7.1.2 Arithmetical Difference Method.- 7.1.3 Fuzzy Set Matching Method.- 7.2 Application to Attribute Types.- 7.2.1 Percent Attribute Type (PAT).- 7.2.2 Number Attribute Type (NAT).- 7.2.3 Discrete Attribute Type (DAT).- 7.2.4 Hierarchy Attribute Type (HAT).- 7.2.5 Multi-Attribute Types (MAT).- 7.3 Matching Example.- 7.4 Clustering.- 8 Profile Initialization and Update.- 8.1 User Profiles Initialization.- 8.1.1 Explicit Profile Specification.- 8.1.2 Specification by Individual User Properties.- 8.1.3 Specification by Representative Program Examples.- 8.1.4 Initialization by Observation.- 8.1.5 Comparison.- 8.1.6 Conclusion.- 8.2 User Profile Reference Table.- 8.2.1 Defmition of User Properties and Attributes.- 8.2.2 Definition of Initial User Profile Attributes.- 8.2.3 Results.- 8.2.4 Exceptions.- 8.2.5 Further Improvements.- 8.3 User Profile Updates.- 8.4 User Groups.- 9 Decision Support System.- 9.1 Characteristic Properties.- 9.1.1 Visitors.- 9.1.2 Program Changes.- 9.1.3 Periodic Events.- 9.1.4 Zapping.- 9.1.5 Sleeping.- 9.2 User Behavior Variables.- 9.3 Reaction to Zapping.- 10 Implementation and Testing.- 10.1 Implementation based on the MHEG Standard.- 10.1.1 User Interface Agent and Presentation Objects.- 10.1.2 MHEG Engine.- 10.1.3 Multimedia Store.- 10.1.4 Management Components.- 10.2 Implementation of the Electronic Program Guide Application.- 10.2.1 MHEG Prototype Implementation.- 10.2.2 HTML/JAVA Prototype Implementation.- 10.2.3 Comparison of MHEG and JAVA Approach.- 10.2.4 Results of Clustering.- 10.3 Evaluation with Respect to Requirements.- 10.3.1 General Remarks.- 10.3.2 Requirements.- 10.4 Summary.- 11 Conclusions.- 11.1 Classification.- 11.2 Suggested Extensions.- 11.2.1 Exploration of New Interests.- 11.2.2 Content Based Retrieval.- 11.2.3 Information Broker.- 11.2.4 Automated TV Program Director.- 11.2.5 Other Information Systems.- Appendix A: Genre Trees.- Appendix B: Personal EPG Application based on JAVA.- Appendix C: The Administration Tool.- Appendix D: Sample Hit List.

O autorze

Hartmut Wittig is directing the corporate strategy department of "Deutsche Telekon Computer Service Management" (DeTeCSM), In the recent years, H. Wittig has directed multimedia and internet software R&D departments and groups. He gave lectures on "Multimedia Systems" in German universities and worked as Editor in Chief of IEE Multimedia Newsletter.

Hartmut Wittig ist Zentralbereichsleiter für Unternehmensstrategie bei der Deutschen Telekom Computer Service Management GmbH (DeTeCSM). Das Buch stützt sich auf seine Erfahrungen in leitenden Tätigkeiten der Internet-Softwareentwicklung. In den vergangenen Jahren war er außerdem als Dozent für Multimedia-Systeme an verschiedenen deutschen Universitäten sowie als Herausgeber der IEEE Multimedia Newsletter tätig.

Blurb

Intelligent agents are rescuer in the information glut. They help users to find information which better corresponds to their interests and needs. This book describes the architecture and basic modules of an intelligent media agent. A personal television guide is described as an example of intelligent help, addressing the problem of managing TV channels by using an intelligent agent.