Scientific Workflow Model for Provenance Analysis Based on Many-sorted High-Level Nets

Scientific workflow is becoming an important paradigm for scientific computations and collaborations. Provenance information plays an important role to reproduce data and prove the soundness of result for scientific workflow. The provenance information can be collected by keeping a complete trace of the workflow execution. In this paper, we give a high level Workflow Net model based on many-sorted signature algebraic nets. The model combines data type and Petri Net, and it is easy to depict the relation between provenance and workflow execution. Based on the model, it is convenient to capture provenance information and make further query.