Enhanced publication

The main motivations behind enhanced publications are to be found in the limits of traditional scientific literature to describe the whole context and outcome of a research activity. The goal is to move beyond the simple PDF (FORCE11 initiative) to support scientists with digital and automated access to the literature and any form of research outcome (e.g. research datasets, ontologies), still without losing the narrative spirit of “the publication” as dissemination means. Enhanced publications are digital objects characterized by an identifier (possibly a persistent identifier) and by descriptive metadata information. The constituent components of an enhanced publication include one mandatory textual narration part (the description of the research) and a set of interconnected sub-parts. Parts may have or not have an identifier and relative metadata descriptions and are connected by semantic relationships.

Background

Scientists are increasingly conducting research using e-Science infrastructures and adopting data driven methodologies. As such they are becoming interested in getting scientific reward for any assets in the research chain, from datasets, software products, devices, to scientific processes. Besides, they would like to be able to re-use the outcome of other’s research to reproduce similar experiments and/or avoid pointless mistakes. Such trends are further leveraged by the fact that also funding agencies and organizations, which are crucial stakeholders in the research chain, are advocating and increasingly making mandatory publishing and citing of any research outcome in order to measure their Return of investment, improve their funding strategies, or gain visibility and scientific rewards. The dissemination of research outcomes via traditional publications, in either paper or digital form, cannot cope alone with these new needs of scientists. To tackle such limits, the new notion of enhanced publication has emerged in several scientific disciplines as a new research result dissemination mean.

Information Systems and Data Models for Enhanced Publications

Existing enhanced publication information systems may significantly vary in terms of their functional goals, typically a combination of:

• Packaging of research outputs
• Web 2.0 reading capabilities
• Interlinking research outputs
• Re-production and assessment of scientific experiments.

In general, such information systems implement data models that describe enhanced publications as consisting of parts, whose purpose is to define one mandatory textual/narrative description of the research conducted (as for the traditional article) and its relationships with other material whose nature and location (local or remote to the system) vary depending on the scientific context (e.g. datasets, images, tables, workflows, devices, services) and the above specified consumption purposes.

Packaging of research outputs

The first enhancement introduced to move beyond the mare digitization of the publication and investigate new avenues in the digital scholarly communication was likely accompanying a digital publication (e.g. PDF file) with supplementary material. In such scenarios, scientists can share packages consisting of publication and supplementary material in an attempt to better deliver hypothesis and results of the research presented in the publication.

Information System	References
Journals with supplementary material policies	Elsevier Supplementary Data policies, SAGE Journals: Author Guide to Supplementary Files
Modular Article	Kircz, Joost G. (2002)
LORE	Gerber, A., & Hunter, J. (2010)

Web 2.0 reading capabilities

This category of approaches focuses on enhanced publication data models whose parts, metadata and relationships are defined with the purpose of improving the end-user experience when visualizing and discovering research materials. These approaches are the natural extension of the traditional publication, oriented to reading, and typically integrate all tools made available by the web infrastructure and its data sources. Specifically, they explore the possibilities of:

1. structuring narrative text into interconnected sub-parts;
2. re-using the universe of web resources to enrich the text;
3. including dynamic forms of content within the text.

Information System	References
D4Science Live documents	Candela, L., Castelli, D., Pagano P., & Simi, M. (2005)
SciVee	Fink & Bourne (2007)
PLOS Neglected Tropical Diseases	Shotton, Portwin, Klyne, & Miles (2009)
The Veteran Tapes Project	van den Heuvel, van Horik, Sanders, Scagliola & Witkamp (2010)
Utopia Documents	Attwood et al. (2010)
Rich Internet Publications	Breure, Voorbij and Hoogerwerf (2011)
SOLE	Pham, Malik, Foster, Di Lauro & Montella (2012)
Elsevier's Article of the Future	Aalbersberg et al. (2012)
Bookshelf	Hoeppner (2013)

Interlinking research outputs

Scientific communities, organizations, and funding agencies are nowadays supporting and welcoming initiatives, standards and best practices for publishing and citing datasets and publications on the web. Examples are DataCite, EPIC and CrossRef, which establish common best practices to assign metadata information and persistent identifiers to datasets and publications. Data publishing and citation practices are today strongly advocated by research communities, which need datasets to become first citizens in the research production chain. Datasets should be discoverable, reusable, and scientists who produce them be scientifically rewarded for their efforts. On this respect, several enhanced publication information systems were built to offer the possibility to enrich a publication with links to relevant research data, in order to strengthen data citation, facilitate data re-use, and reward the precious work underlying data management procedures. In these context an enhanced publication consists of a narrative description (embedded or linked to) and relationships to external references. In this sense, references survive together (in the scope) of the publication.

Re-production and assessment of scientific experiments

How can research results be effectively “peer-reviewed” or more generally evaluated for their quality? The creation of enhanced publications is certainly an important step ahead in this direction, since researchers can count on further information contextual to the experimentation described in the research paper. However, even the acquisition of research data used in the experimentation is often not enough in many research fields to assess quality of results. Indeed, scientists and reviewers should be equipped with the tools necessary to repeat the experiment, and the authors should share such tools as part of their enhanced publication. In fact, this does not sound as a surprise in e-Science infrastructures, where researchers are urging for tools to disseminate and reuse the whole context of their research. To address such demands, several solutions were proposed in the literature on how to share executable components via enhanced publications. In enhanced publications with executable parts, narrative parts are accompanied by executable workflows with the purpose of reproducing experiments.