Prof' Dr' Bettina Berendt

1
Web Usage Mining Evaluation

• Prof. Dr. Bettina Berendt
• Humboldt Univ. Berlin, Germany
• www.berendt.de

2
Stages of knowledge discoverydiscussed in this
lecture
Application understanding
3
The slides (with more context)

• This lecture is based on the slides of the
  tutorial
• Berendt, B., Spiliopoulou, M., Menasalvas, E.
  (2004). Evaluation in Web Mining. Tutorial at the
  15th European Conference on Machine Learning /
  8th European Conference on Principles and
  Practice of Knowledge Discovery in Databases
  (ECML/PKDD'04), Pisa, Italy, 20 September 2004.
• http//www.wiwi.hu-berlin.de/berendt/evaluation0
  4/, Parts I III
• This lecture is structured by the CRISP-DM
  process, with the following corresponding pages
  in the tutorial
• Introduction Tutorial Part I / Motivation and
  Background
• Evaluation and CRISP-DM stage Business/applicatio
  n understanding Part II / The CRISP-DM
  Reference Model
• Evaluation and CRISP-DM stage Modeling pp. 10,
  97
• Evaluation and CRISP-DM stage Evaluation Part
  III
• Evaluation and the whole CRISP-DM process Part
  II
• Additional remark on best practices / standard
  workflows for Web mining (in the tutorial slides
  CATs for Web Mining) look at WEKAs
  KnowledgeFlows

4
Thank you!