With the proliferation of heterogeneous wireless networks, classical information-theoretical methods can no longer provide accurate mathematical models to analyze the complicated network performance for the regulation and optimization of networks. The emerging landscape of big data and machine learning has provided a new paradigm for the design and optimization of intelligent wireless networks. In this article, a general architecture for data-cognition-empowered intelligent wireless networks is proposed. Characteristics of wireless data, the frequently used utilities for data cognition, and data cognition methods are discussed to depict the building blocks and challenges in the exploration of the architecture for data-cognition-empowered intelligent wireless networks.