The behaviour of multiple stock markets can be described within the framework of complex dynamic systems. A representative technique of the framework is the dynamic interaction network (DIN), recently developed in the bioinformatics domain. DINs are capable of modelling dynamic interactions between genes and predicting their future expressions. In this paper, we adopt a DIN approach to extract and model interactions between stock markets. The network is further able to learn online and updates incrementally with the unfolding of the stock market time-series. The approach is applied to a case study involving 10 market indexes in the Asia Pacific region. The results show that the DIN model reveals important and complex dynamic relationships between stock markets, demonstrating the ability of complex dynamic systems approaches to go beyond the scope of traditional statistical methods.