Repurposing Open Source Tools for Open Science: a Practical Guide
Open science – i.e., the practice of conducting scientific research in a way that is transparent, reproducible, and openly accessible to all – is gaining increasing attention and popularity. For instance, the European OpenAIRE project is supporting a vision for open access and open data, by providing researchers with the necessary guidance and tools to conduct open science. Moreover, YERUN, a network of Young European Research Universities founded a working group to coordinate the adoption of open science research practices at European universities. This poster demonstrates how open science can be further advanced using existing tools that are already widely used by open source communities. We combine systems originating from (1) the open source movement, (2) the digital legal initiative, and (3) recent results from the open science movement itself. Our idea is to dedicate one service to each step in the research process. For example, GitHub can be used for document management. OriginStamp, a blockchain-based distributed timestamping service, can be used to prove the originality of data and computations. Travis, a continuous integration service, can be used for automation and objective execution and, lastly, the research data repository Zenodo can be used to generate immutable, durable, and accessible final research artifacts. However, only in combining each of the services used during the individual steps of the research cycle, will scientists be able to generate reproducible and open scientific results. In the vision we present, the blockchain technology that enables the tamper-proof timestamping of final research artefacts and – even more importantly – intermediate results, is a game changer. The existence of tamper-proof documentation for each incremental research step improves reproducibility without forcing researchers to perform each research step immediately open to the public.
In this poster, we exemplify the proposed combined approach using a typical research scenario: We describe how to collaboratively write a paper in LaTeX and publish it in an open research data repository. We also provide background information on the open source services used in each step. Finally, we outline how researchers can adapt the presented seven-step procedure to other scientific tasks, such as planning a user study, running a simulation, or developing a software prototype.