The massive data output of many scientific disciplines nowadays calls for standardized approaches to the preservation of data, experimental methods, and metadata associated to the scientific discovery process. These approaches are more and more frequently embedded in so-called scientific workflows, which are proving to be critical elements in data-intensive science. Amongst other benefits for experimental sciences, scientific workflows support the automation of scientific methods and the incorporation of best practices in scientific data management, which result in extended capabilities for validating and reproducing experimental outcomes and encourage new ways of scholarly publication.”.

During the last three years, the FP7 project Wf4Ever led by the Spanish SME iSOCO, with partners University of Manchester, Universidad Politécnica de Madrid, Poznan Supercomputing and Networking Center, University of Oxford, Instituto de Astrofísica de Andalucía and Leiden University Medical Center, has researched new methods, models, and tools to make this vision possible, with special focus on the preservation and conservation of experimental knowledge. And many have been the challenges faced in this quest. Exclusively relying on workflow specifications does not suffice to guarantee reusability, shareability, reproducibility, or a better understanding of scientific methods. Over time, experimental environments evolve and some disappear, the services and tools used may change or evolve too, the data used in the experiment may be updated or no longer accessible or available, even teams of scientists change and knowledge can be lost, and satisfying the collaboration and information access needs of the diverse stakeholders related to scientific development, including scientists, librarians and publishers, requires thinking out of the box.

To overcome these challenges, Wf4Ever has proposed the concept of Research Objects as the means to encapsulate all the information and materials related to a scientific experiment in a single information unit. Research objects may include workflows and annotations aimed at humans and computers, which describe the operations performed by a workflow; provide details like authors, versions, and citations; and link to other resources, such as the provenance of the results obtained by running the experiment or the datasets used as input. Research objects therefore provide a comprehensive view of the experiment, becoming carriers of experimental knowledge. As a consequence, research objects enable the exchange of such knowledge between teams of scientists, the review and validation by community members of experimental outcomes and claims, the training and education of new members of the scientific communities, and the fight against decay both in the experiments and the methods enclosed.

Wf4Ever has proved to be an extremely dynamic, prolific, and impactful project, with key outcomes including means to represent, manipulate, conserve, search, and recommend research objects as well as methods for their optimal use and conservation in scientific disciplines. The Research Object model is at the core of these contributions, enabling the representation of research objects and supporting a landscape of interoperable tools for research object management. Among these tools, the well-known scientific platform myExperiment, extended during Wf4Ever with research object capabilities, and the newly created Research Object Hub provide a variety of research object services and functionalities to a broad range of users, from scientists to librarians and publishers. Wf4Ever principles have also reached out to a broad spectrum of scientific communities, ranging between Astrophysics and Genomics to Music, through a collection of best practices and plans that provide users with guidelines for a systematic approach to manage scientific knowledge based on research objects. Additionally, lively business sectors in the content delivery domain like media and publishing are witnessing an increasing number of entrepreneurial initiatives around the notion of research objects as content aggregations.

The Wf4Ever project has now finished after three years of intense work but its legacy continues. The lessons learnt, experiences and knowledge about research objects produced during Wf4Ever have inspired the creation of ResearchObject.org as a knowledge hub for the research object community. The portal aims to disseminate knowledge about research objects, the underlying concepts, continuous uptake, and the most recent developments in the area, including new insights and viewpoints, to a broader community beyond the Wf4Ever consortium. Amongst such objectives, it is worth noticing the promotion of a broader vision of research objects as general-purpose containers of research information, which extends the notion of workflow-centric research objects. Paraphrasing Winston Churchill: Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning… for research objects

Print