Abstract:Parametric product catalogs provide unique context for data mining techniques. The specific user access mode is analyzed. The availability of structured content data is taken into consideration. Three mining scenarios (category association, feature evaluation and product recommendation) are discussed. Related mining techniques are introduced and adjusted to the context. A system framework is put forward to show how these techniques can support each other efficiently to build a more integrated and intelligent system.

Data mining in parametric product catalogs