[Objective] This paper proposes a data visualization model to retrieve, analyze and present historical records from a data coalition, aiming to improve the knowledge discovery. [Methods] We constructed a model for the visual data analysis system, and then used a big data platform to examine its feasibility. [Results] The proposed system could analyze massive historical data and then support the decision making procedures. [Limitations] The current visual analysis result views could be further improved by adding more chart templates. [Conclusions] The proposed system could analyze historical data from the library alliance and provide valuable information for decision makers.