Context-aware integration of data-centered services

Data as a Service (DaaS) emerges as a new trend for exchanging data between independent data owners and data users so that data can be acquired on demand through standard protocols across heterogeneous platforms. It is usually a user-interactive and iterative process to compose the services into various data-driven business scenarios of data acquisition, analysis, and other processing activities. Context information can facilitate the recommendation and selection of the services based on the runtime service behavior, user preferences, and historical results. The paper proposes a context-aware approach for integrating data in dynamic changing business processes following the DaaS architecture, as an extension of our previous work on the service-based framework for pharmocogenomics data integration. A context model is defined in three dimensions including the tasks, the services, and the users, from both static and dynamic perspectives. The monitoring and feedback mechanism is introduced to enable continuous profiling of the runtime behavior and updating of the context properties. The paper takes the biomedical research as the motivating scenario to illustrate and evaluate the effectiveness of the proposed approach.