Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining

Abstract

KDD-2010, the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, is being held in Washington, DC, USA, on July 24--28, 2010. KDD is the leading international forum for the exchange of research results and practical experience in the field of knowledge discovery and data mining. As the quantity of data available to organizations and individuals continues to grow rapidly, and the need to extract useful knowledge from them becomes more intense, scientists, government workers and business people turn to the KDD community for solutions. This volume contains a snapshot of a year of developments in this field; we hope you will find it useful and rewarding. The KDD-2010 technical program features four parallel research tracks and an industrial / government track. The program also features keynotes from leading creators and consumers of KDD technology, 12 workshops, 12 tutorials and one panel. The 2010 KDD Cup competition focuses on educational data mining to support improvements in the field of computer aided instruction. Dozens of technical demonstrations and exhibits from vendors and other organizations underscore the conference's dual role as the leading industry and academic forum to discuss the advances in this field of research. The call for papers attracted 578 research papers and 101 industrial and government submissions from around the world. Each paper was independently reviewed by three members of the program committee for originality, significance, technical quality, and clarity of presentation. This year's research track introduced an author-feedback phase in the review process, in which authors were invited to comment on the preliminary reviews that they received. The objective of the feedback phase is to ensure greater transparency and fairness, as the authors' responses are taken into account in a subsequent discussion phase moderated by Senior Program Committee (SPC) members. There was much discussion among the reviewers in the subsequent discussion phase before the final decisions. In the end, the program committee accepted 77 papers for long presentations and 24 papers for short presentations into the research track, representing an aggregated acceptance rate of 17.4%. This year's Industry and Government track emphasized the successful uses of KDD technology, including deployed applications incorporating KDD technologies and discoveries of valid, novel, understandable, and demonstrably useful patterns from large datasets in industry and government, as well as emerging applications and technology, including challenges and issues arising from attempts to deploy KDD technology to solve specific industry or government problems. The industry and government track of the conference accepted 11 papers for long presentations and 9 papers for short presentations into the program, representing an aggregated acceptance rate of 19.8%. We are glad to see that the conference remains strongly competitive and of very high quality. The resulting program was notable for its diversity and vitality. Alongside traditional KDD topics like classification, clustering, frequent sets, and temporal data, we also saw papers in rapidly growing areas like social network and graph mining, privacy, recommendation, ranking, topic modeling and transfer learning. Application areas included the Web, bio-informatics, mobility data, healthcare, marketing, and many others.