Creating Interactive Scientific Publications using Bindings

Abstract

Many scientific publications report on computational results based on code and data, but even when code and data are published, the main text is usually provided in a separate, traditional format such as PDF. Since code, data, and text are not linked on a deep level, it is difficult for readers and reviewers to understand and retrace how the authors achieved a specific result that is reported in the main text, e.g. a figure, table, or number. In addition, to make use of new the opportunities afforded by data and code availability, such as re-running analyses with changed parameters, considerable effort is required. In order to overcome this issue and to enable more interactive publications that support scientists in more deeply exploring the reported results, we present the concept, implementation, and initial evaluation of bindings. A binding describes which data subsets, code lines, and parameters produce a specific result that is reported in the main text (e.g. as a figure, table, or number). Based on a prototypical implementation of these bindings, we propose a toolkit for authors to easily create interactive figures by connecting specific UI widgets (e.g. a slider) to parameters. In addition to inspecting code and data, readers can then manipulate the parameter and see how the results change. We evaluated the approach by applying it to a set of existing articles. The results provide initial evidence that the concept is feasible and applicable to many papers with moderate effort.

The role of metadata in reproducible computational research

“…The concept of bindings connects interactive parts of an ERC workflow with the underlying code and data. 170 Instead of trying to establish a common standard and single point for metadata, the ERC intentionally skips formal metadata and exports the known information into multiple output files and formats, such as Zenodo metadata as JSON or Datacite as XML, accepting duplication for the chance to provide usable information in the long term.…”

“…While Schema.org tags have been used for indexing data, to our knowledge there is no journal that supports, much less encourages, semantic markup of specific terms within a manuscript. There has been tacit support for such inline markup in newer manuscript composition tools such as Manubot 236 or the ERC's bindings, 170 but generally such terms could disambiguate concepts, point to the provenance of…”

Publishing computational research - a review of infrastructures for reproducible and transparent scholarly communication

“…o2r [22] originates from the geosciences and addresses publishers who want to extend their infrastructure via a reproducibility service during the process of paper submission [28]. Authors can create interactive figures, allowing readers to change model parameters using a slider [29].…”

“…Moreover, identifying specific parameters buried in the code and finding out how to change these can be a daunting task. The concept of nano-publications [49] or bindings [29] might help to solve these issues. A further need in this context is a UI for comparing original and manipulated figures, since differences in the figure after changing parameters might be difficult to spot.…”

Publishing computational research – A review of infrastructures for reproducible and transparent scholarly communication

“…o2r (Nüst et al 2017) addresses publishers who want to extend their existing infrastructure by a reproducibility service during the process of paper submission (Nüst 2018). Authors can also create interactive figures, allowing reviewers and readers to check the robustness of the results, e.g., by changing model parameters using a slider (Konkol, Kray, and Suleiman 2019). REANA (Šimko et al 2019;Chen et al 2018) provides a formal specification to guide authors through the process of capturing input datasets, code, and the computational environment.…”