A new dominant innovation framework was proposed to address the shortcomings of the design knowledge and single-source data-driven traditional requirement mining methods, such as over-reliance on design experience, incomplete requirement mining, and lack of requirement forecasting. The unstructured text data was used as the object to extract and classify the features of requirement from the users’ perspective. The operational big data of product was used as the object to mine the users’ requirement from the product’s perspective. The requirement prediction theory was applied to predict the implicit requirement of future products and construct a data-driven dominant innovation matrix. Compared with the traditional knowledge-driven market research and the single-source data requirement mining methods, the proposed framework was more objective and comprehensive for explicit requirement mining, while taking into account the future implicit requirement mining. The mini-printer of an intelligent manufacturing enterprise was used as a case to verify the effectiveness of the proposed framework.