Abstract: Scientific data often come from networks with complex relationships between their entities and can be properly modeled as semantic graphs. However, once designed, there is no simple way to cross through different designs in graph databases. The goal of this research is to specify and implement a framework to overcome these limitations, allowing users to build and explore arbitrary perspectives in graphs. The framework uses the concept of views to represent a perspective. The main contribution is to help scientists run models and analyze network (graph) data according to their specific design needs. The framework is under implementation and validation using a case study on water resource data Ver menos