Seminar: ECML/PKDD, Warszawa, Industrial Conference on Data Mining, Leipzig, Personalized Web Group (Pewe) Bratislava, Coping With Semantics In Web Usage Mining

A report from long-term visit.

Semantic enrichment of clickstreams (web visitor paths) is a distinguished research trend in web usage mining; the availability of such data can make a large qualitative difference to the results of data mining. The talk will address two problems related to this topic: a) obtaining semantic data and ensuring its up-to-dateness is usually expensive b) the semantic data tend to elarge the attribute space which may lead to the curse of dimensionality. The proposed solution for practical and timely acquisition of semantics is a plug-in for widely used web analytics software – Google Analytics. A second part of the talk will cover a work-in-progress heuristic for dimensionality reduction of semantically enriched clickstreams.