Multimedia Data Mining in Conjunction with Acm Sigkdd Eighth International Conference on Knowledge Discovery and Data Mining Table of Contents Multimedia Data Mining Using P-trees Scale Space Exploration for Mining Image Information Content Object Boundary Detection for Ontology-based Image Classifi

Foreword Since the beginning of the century there have been two successful international workshops on multimedia data mining at the KDD forums: MDM/KDD2000 and MDM/KDD2001, in conjunction with KDD2000 (in Boston) and KDD2001 (in San Francisco), respectively. These workshops brought together numerous experts in spatial data analysis, digital media, multimedia information retrieval, state-of-art data mining and knowledge discovery in multimedia database systems, analysis of data in collaborative virtual environments. For more information about the workshops see the reports on the workshops in SIGKDD Explorations (2 (2), pp. 103-105 and 3 (2), pp. 65-67, respectively). Participants in both workshops were pleased with the event and there was consensus about the necessity of turning it into an annual meeting, where researchers, both from the academia and industry can exchange and compare both relatively mature and green house theories, methodologies, algorithms and frameworks for multimedia data mining. This workshop is organized in response to this interest. Being a third edition, the workshop this year is aiming to create a stimulating atmosphere for discussing the theoretical foundations of multimedia data mining, frameworks, methods and algorithms for integrated pattern extraction from multimedia data, multimedia data preprocessing, novel architectures for multimedia data mining, and applications of multimedia data mining in different areas. Consequently, the papers selected for presentation at the Third International Workshop on Multimedia Data Mining (MDM/KDD'2002) held in conjunction with the 7th and Applications of Multimedia Data Mining (with two subgroups of applications: in medical image analysis and in content-based multimedia processing). This grouping bears some similarity with the last year workshop, where there was similar emphasis on the research in the area of frameworks and methodologies, and on the research in the application area. The works selected for presentation at this workshop form more cohesive body of work, which indicates that the field has made a step forward towards achieving some level of maturity. As part of the SIGKDD conference series the workshop follows a rigid peer-review and paper selection process. Once again, we would like to thank all those, who supported this year's efforts on all stages – from the development and submission of the workshop proposal to the preparation of the final program and proceedings. We would like to thank all those who submitted their work to the workshop. In a good data mining tradition, a pattern is emerging – as in the previous workshop there were submissions from 10 different …