One of the most dynamically developed areas of the research in engineering is definition and application of knowledge in product model. Product model is in the centre of product lifecycle management systems to support engineering activities with comprehensive representation of product information. Recently, knowledge is defined in the form of active features in the product model. When an active knowledge feature or value of its parameter changes, other feature parameters those are in contextual connection with it change according to its content. In this paper, a new modeling method is introduced to elevate active knowledge from the product feature level to the product behavior level. This allows for coordination of product definition on the behavior level. Result of a recent study on the currently applied product definition is utilized at the establishment of a connection between the proposed higher level knowledge features and the currently used active knowledge features. In the proposed modeling method, result of human intent analysis is applied at situation based generation of features in the product structure. The new method of product feature definition and the demanded new features are explained in this paper. In order to easy implementation of the proposed modeling in large modeling systems, the method of the feature driven product definition is the same as in the current leading industrial product lifecycle management systems.