Workshop at the Third International Conference on Knowledge Discovery and Data Mining
Sponsored by the Association for the Advancement of Artificial Intelligence
The KDD 1997 Workshop on Issues in the Integration of Data Mining and Data Visualization was held on Sunday, August 17, 1997 in Newport Beach, California. The workshop description is listed below:
Data visualization deals with the effective portrayal of data with a goal towards insight about the data. Typically, the data is of high volume, multidimensional in nature, and does not lend itself to easy display. The data is also often non-spatial and temporal in nature.
Data visualization software systems are very popular with end-user domain scientists who require visual tools to explore and analyze their data. These visual tools however are used strictly as output of the exploration process and have received much attention whereas the input issues to the exploration process still have not. The KDD community is concerned with two aspects of visualization techniques: 1. its use at the "back-end" of the exploration process to help understand models extracted by data mining algorithms, and 2. scalability issues in visualization: how do we make it efficient in presence context of large databases where data access is expensive. The visualization community looks at KDD and analytic methods also as applications to generate displays. However visualization can be used as input to KDD and analytic tools; it can also be used to support computational steering. An effective visualization front-end can guide a data mining algorithm in its search and may result in much better and more easily acceptable solutions. This workshop will continue the discussions started at the first two workshops and focus on these and other issues that make a case for integrating KDD and visualization technologies.
Two previous workshops (Siggraph '90 and Visualization '91) have dealt with areas such as high-level requirements for data structures and access software, and data visualization environments. The first and second workshop on Database Issues for Data Visualization were held in 1993 and 1995 and explored the fundamental issues. A number of experimental, prototype, and research systems were presented. The second workshop also saw a beginning interest with data mining and visualization integration. This trend, so significant in the commercial sector today, is in its infancy and is in need of much research attention.
Position statements and papers were welcome on the following issues as they relate to KDD and data visualization integration. The organizers wanted to keep discussions focused on the end result, which is improving the integration of data mining and knowledge discovery systems with visualization:
- Requirements Visualization places on Knowledge Discovery Systems
- Data Models and Access Structures
- Modeling the User - Tasks, Processes, Support Issues
- Advanced User Interfaces for Data Mining
- Visual Languages for Data Mining
- System Integration Issues
- Computational Steering for Data Mining
- Scalability to Large Databases
- Distributed, Heterogeneous Data Set Issues - Data and Computation Sharing
- Examples of Integrated Systems
- Applications of Integrated Systems
- Georges Grinstein
- Institute for Visualization and Perception Research
- University of Massachusetts at Lowell
- Lowell, MA 01854, USA
- Andreas Wierse
- Institute for Computer Applications
- Dep. Computersimulation and Visualization
- Pfaffenwaldring 27
- D-70550 Stuttgart, Germany
- Usama Fayyad
- Microsoft Research
- Redmond, WA 98052-6399, USA
Workshop Technical Reports
Some KDD workshops are available as technical reports. For contents and ordering information, consult the AAAI Press catalog.