This paper deals with recent results of a long term research program in human intent driven and knowledge based product modeling. The aim of the research was grounding connection between situation and event based generation of product features in current industrial product models and the content based model extension in an earlier result by the authors. In this paper, research is focused on knowledge feature definition for adaptive feature control. The authors considered their Virtual Engineering Space concept as the utmost objective in automation of engineering activities. Following this, knowledge definition methods are analyzed, contextual connections of active knowledge definitions in the context based product model extension are placed in the communication, and control of product features by adaptive actions is discussed.