|
- Panel Chair Presentation (PDF, 14.4 KB) - Panel Chair Position Statement (PDF, 35.7 KB) - CPExchange Presentation (PDF, 200 KB) -
CRISP-DM
Position Statement (PDF, 14.8 KB) |
KDD-2000 Sixth ACM SIGKDD International Conference on KDD Process Standards Panel
Chair: Ismail Parsa (Epsilon) Panel Chair Presentation (PDF, 14.4 KB) Panel Chair Position Statement (PDF, 35.7 KB) Adoption of universally acceptable and sustainable "standards" -- in terms of terminology, evaluation and methodology -- is key to KDD's continuing success. Today, the process of knowledge discovery and data mining is far from automated (i.e., still requires human interaction) and, therefore, is difficult to deploy effectively. Many commercial tools lack functionality and scalability that customers require. There are institutions/ on-going efforts working independently toward streamlining the whole or part of the KDD and/or related processes. Among them are CPEX, CRISP-DM, DMG, MDC, OLEDB DM and TASF.
This panel bring together representatives from these institutions/ on-going efforts with the objective of informing the KDD community on the current state of their work and, possibly opening the doors for mutual collaboration.
RELATED PROCESSES (in alphabetical order of occurrence)
- CPEX (Customer Profile Exchange) offers a vendor-neutral, XML-based open standard for facilitating the privacy-enabled interchange of customer information across disparate enterprise applications and systems. [www.cpex.org] CPExchange Presentatation (PDF, 200 KB) Participant from CPEX: Brad Husick.
- CRISP-DM is an industry consortium developing an industry-neutral and tool-neutral CRoss-Industry Standard Process Model for Data Mining. [www.crisp-dm.org] CRISP-DM Position Statement (PDF, 14.8 KB) Participant from CRISP-DM: Ruediger Wirth.
- DMG, the Data Mining Group, is a consortium of industry and academics formed to create standards, starting with PMML, (XML-based) for defining and sharing predictive models. [www.dmg.org] Participant from DMG: Robert Grossman.
- OLEDB DM (OLE DB for Data Mining), a Microsoft effort extending SQL databases through a new API to better support data mining operations. [www.microsoft.com/data/oledb/dm.htm] Participant from OLEDBDM: Usama Fayyad.
- TASF (The Analytic Solutions Forum), an industry consortium whose mission is to establish solution-oriented performance criteria and interoperability requirements within and between classes of decision support tools (including but not limited to OLAP or On-line Analytical Processing, data mining, data visualization, text processing, and decision analysis). [www.tasf.org] Participant from TASF: Erik Thomsen. |
|
