A Python Library for Provenance Recording and Querying

Abstract

In many application domains the provenance of data plays an important role. It is often required to get store detailed information of the underlying processes that led to the data (e.g., results of numerical simulations) for the purpose of documentation or checking the process for compliance to applicable regulations. Especially in science and engineering more and more applications are being developed in Python, which is used either for development of the whole application or as a glue language for coordinating codes written in other programming languages. To easily integrate provenance recording into applications developed in Python, a provenance client library with a suitable Python API is useful. In this paper we present such a Python client library for recording and querying provenance information. We show an exemplary application, explain the overall architecture of the library, and give some details on the technologies used for the implementation.