As a mega-trend in IT, services computing provides different types of services for widely distributed data sharing and aggregation that span heterogeneous networks and domains. With these services, the need to visualize massive data is growing since there are a large number of traditional applications and it is quite common for those applications to invoke the services. Unfortunately, the visualizations in existing systems do not satisfactorily resolve the basic dilemma of being readable both for the global structure of a large service repository and also for detailed data analysis of local service interaction. To address this problem, we analyze the characteristics of data from Web service and present a hybrid data representation that combines the advantages of two representations: Parallel Coordinates are used to show the global structure of a whole repository, while arbitrary portions of the repository can be shown by NodeTrix to better support the analysis of data. A software prototype is described, which we can use to visualize for service browsing and interacting, such as side-by-side overview and detail views, visualization of fast data searching, and smooth transitions between Parallel Coordinates and NodeTrix.