HDM: A Client/Server/Engine Architecture for Real Time Web Usage Mining

Abstract : The behavior of the users of a website may change so quickly that it becomes a real challenge to attempt to make predictions according to the frequent patterns coming from the analysis of an access log file. In order to reduce the obsolescence of behavioral patterns as much as possible, the ideal method would provide frequent patterns in real time, making the result immediately available. In this paper, we propose a method for finding frequent behavioral patterns in real time, whatever the number of connected users. Considering how fast frequent behavior patterns may have changed since the time the access log file was analyzed, this result thus provides completely appropriate navigation schemata for predicting user behavior. Based on a distributed heuristic, our method also tackles and provides answers to several problems within the framework of data mining: the discovery of ‘interesting zones’ (a large number of frequent patterns concentrated over a period of time, or ‘super-frequent’ patterns), discovering very long sequential patterns and interactive data mining (‘on-the-fly’ modification of the minimum support).
Document type :
Journal articles
Complete list of metadatas

https://hal-lirmm.ccsd.cnrs.fr/lirmm-00191952
Contributor : Christine Carvalho de Matos <>
Submitted on : Monday, November 26, 2007 - 11:43:38 AM
Last modification on : Saturday, February 23, 2019 - 7:06:02 PM
Long-term archiving on : Monday, April 12, 2010 - 5:05:23 AM

File

D154.PDF
Files produced by the author(s)

Identifiers

Collections

Citation

Florent Masseglia, Maguelonne Teisseire, Pascal Poncelet. HDM: A Client/Server/Engine Architecture for Real Time Web Usage Mining. Knowledge and Information Systems (KAIS), Springer, 2003, 5 (4), pp.439-465. ⟨10.1007/s10115-003-0097-6⟩. ⟨lirmm-00191952⟩

Share

Metrics

Record views

211

Files downloads

245