With the advent of knowledge economy, the strategic gravity of ‘really new product/service’ becomes more highlighted in competitive market. Among others, the creative ideation triggers the process of new product/service development. Thus far, however, idea generation has mostly hinged on qualitative brainstorming-based approaches. These intuitive tools are still useful but subject to critical shortcomings in that the potential domain of knowledge base is limited. The drawback becomes an issue as a deluge of data is accumulated in the knowledge domain. The data, either textual or quantitative, may be exploited as the valuable source of new idea. Taking that point account, this research proposes a conceptual framework for data-based idea generation. Overall, the framework is comprised of two major stages. The first stage deals with the construction of structured data base. Here, the spectrum of raw data may be diverse and unstructured: customer review data, patent data, business model data, app-store data, to name a few. The second stage proposes some creativity algorithms and generates new ideas. Specifically, the three types of creativity algorithms are developed and applied: analytical algorithm, combinative algorithm, and visualization algorithm. These individual algorithms may be collectively and flexibly utilized. By doing so, latent information is extracted and converted into visual/innovative knowledge for ideation.