Abstract: Big data context plays a facilitating or constraining role for enterprises to effectively develop and utilize big data-related resources. However, its conceptual definition and dimensional structure are vague, and its impact on specific behaviors is not fully recognized and valued. Rooted in the big data application practices of digital pioneer retail enterprises, this paper defines and constructs the concept and dimensional structure of big data context. It is found that big data context contains two levels: Omnibus context and Discrete context. The Omnibus context is the leading development environment created by the enterprise in the process of promoting the application of big data technology, and its basic elements include executive data awareness, data strategy, data life cycle and data value. The Discrete context is a more micro and specific operational environment that directly affects employees’ actual work behaviors and attitudes, and its basic elements include data attributes, big data organizational image and big data organizational climate. The Omnibus context is conducive to shaping the Discrete context and can be regarded as the antecedent variable of the Discrete context. The Discrete context supports the Omnibus context and mediates the action of the Omnibus context on employee behavior. This theoretical framework enriches and deepens the perception of big data context and makes the definition of context, which is more broadly or vaguely treated, clearer.

Key words: big data context, dimensional structure, grounded theory, concept definition