Session Chairs:

Gautam Das
Microsoft Research
One Microsoft Way
Redmond WA 98052
gautamd@microsoft.com

Mohammed J. Zaki
Computer Science Dept.
Rensselaer Polytechnic Institute
Troy, NY 12180
zaki.AT.cs.rpi.edu

Large-Scale Data Mining

In the present times we are witnessing an explosive growth in the amount of data that is being collected in the business and scientific arena. Data warehouses are filling up with huge amounts of data in every conceivable form. In most cases, the sheer size of the datasets prevents the vast majority of the data from being deeply analyzed. Often large portions may not have been examined at all. The field of Data Mining (or Knowledge Discovery in Databases) attempts to develop automatic procedures that search these enormous data sets to obtain useful information that would otherwise remain undiscovered. Such new knowledge can take the form of patterns, rules, clusters, or anomalies that exist in the massive datasets. These discoveries could potentially be of great significance to scientific or business organizations. Given the size and dimensionality of the datasets, high performance algorithms and systems are an integral component of a successful data mining solution.

The objective of this special session is to bring together technologists and researchers at the forefront of this exciting field to present and discuss their state-of-the-art work. Authors are invited to submit original unpublished manuscripts for the special session on Large Scale Data Mining.

Topics of interest include (but are not limited to):

Efficient, scalable, sequential or parallel and distributed algorithms for various data mining techniques, such as

association rules
classification
clustering
sequence mining

Studying the design of fast methods for the overall data mining process, from the initial data selection to the extraction and management of discovered knowledge.
Scalable data mining on heterogeneous data sources (e.g. the Web, sequence data, images, video, etc).
Development of scalable data mining systems in e-commerce, retail, finance, the sciences, etc.

Accepted Papers:

A scalable approach to balanced, high-dimensional clustering of market baskets, A. Strehl and J. Shosh, Univ. of Texas at Austin

Dynamic integration of decision committees, A. Tsymbal, Univ. of Jyvaskyla

Incremental mining of constrained associations, S. Thomas and S. Chakravarthy, Univ. of Texas at Arlington

Scalable, disctibuted and dynamic mining of association rules, V.S. Ananthanarayana, D.K. Subramanian, and M. NarasimhaMurty, Indian Institute of Science, Bangalore

Papers published as regular papers in the HiPC'00 Conference Proceedings (Springer-Verlag).

Submission Guidelines:

The paper must be clearly identified as submitted to "Large-Scale Data Mining" session. Other submission guidelines are identical to the HiPC guidelines. Papers are to be sent to the HiPC program Chair. The guidelines are summarized here (for details, see www.hipc.org). Submit original research papers not to exceed 15 double-spaced pages of text using 12-point size type on 8.5 x 11 inch pages. Figures and Tables may use additional pages. Preferably send your paper as a correct PostScript (level 2) file. Ensure the PostScript prints on PostScript printers using 8.5x11 paper. In addition to the PostScript, your Email must include, in ASCII form: title, author name(s), abstract, postal address, e-mail address, and telephone and fax numbers. Include "Large-Scale data Mining" in the ASCII header as well as in the paper title page.

Send Electronic submissions to: hipc2000@ac.upc.es

Alternatively send 6 hard copies (by mail, not fax) to the Program Chair at the address:

Mateo Valero
Dept. de Arquitectura de Computadores
Universidad Politecnica de Catalunya
c/ Jordi Girona 1-3, Modulo D6
08034 Barcelona, SPAIN
Email: mateo@ac.upc.es

Important Dates:

Papers Due: May 1st, 2000
Acceptance Notification: June 30th, 2000
Camera Ready Papers Due: August 15th, 2000

Number of Visitors