Skip to Main content Skip to Navigation
Master Thesis

Towards user assistance in Data Mining

Abstract : The need for a unifying and comprehensive methodology is considered one of the major challenges to Data Mining. Indeed, the existing methodologies and Data Mining tools say what to do, but not how. The user is thus faced with many choices and no help. In the best situation, we can find "local" support for a sub-task, for a sub-problem, etc., without any interaction with previous and next steps in the process. There is therefore no global vision over the development of the process and in the worst case many projects fail (due to lack of resources, their complexity) or the results are suboptimal. The Data Mining software is also increasingly complex and the need for assistance for advanced users amplifies. The "general public" solutions (tools for analyzing data on Smartphones, for example) strengthens the interest of thinking about software based on methodologies for guiding users. It is proposed first to describe an architecture that provides the appropriate environment for the assistant and then to describe the chain of possible tasks during the Business Understanding and Data Understanding steps of CRISP-DM process. The assistance is provided taking also in account the past similar situations. The proposal for the conceptualization of these two steps will be exemplified through a case study.
Complete list of metadata

Cited literature [50 references]  Display  Hide  Download
Contributor : Co-Responsables Du Mri V. Gouranton Et S. Blazy Connect in order to contact the contributor
Submitted on : Friday, October 28, 2011 - 11:29:34 AM
Last modification on : Tuesday, October 5, 2021 - 10:18:03 AM
Long-term archiving on: : Monday, January 30, 2012 - 11:14:38 AM


  • HAL Id : dumas-00636764, version 1



Cristina Oprean. Towards user assistance in Data Mining. Databases [cs.DB]. 2011. ⟨dumas-00636764⟩



Record views


Files downloads