The transformation from human- to machine-readability of published content is a key feature of the dramatic changes experienced by academic publishing in the last decade. Non-machine readable PDFs, either digitally born or scanned from paper prints, require significant additional effort of post-publication markup and data extraction into a structured form, in order to address issues of interoperability and reuse of publications and data (Agosti 2006, Penev et al. 2010, Agosti 2016). A partial solution to the problem is the pre-publication markup which can be generic (e.g., for the article metadata and the standard division into article sections such as Introduction, Material and Methods and others) and domain-specific (e.g. markup of taxon names or biological collection codes). The open access journal ZooKeys was the first to implement both generic and domain-specific markup which was adopted thereafter by PhytoKeys, MycoKeys, Journal of Hymenoptera Research, Deutsche Entomologische Zeitschrift, Zoosystematics and Evolution and other Pensoft journals (Penev et al. 2010, Penev et al. 2012). The domain-specific, pre-publication markup was possible thanks to the TaxPub XML schema, developed by Plazi and later endorsed as an extension to the Journal Archival Tag Suite (JATS) by the National Library of Medicie of the USA (Catapano 2010). The pre-publication markup required creation of some tools to facilitate the process (for example, Pensoft Markup Tool and Pensoft Wiki Convertor) and also other tools to visualise the results of it (for example, Pensoft Taxon Profile, or PTP).

The next stage of development of integrated narrative and data publishing was landmarked by the Biodiversity Data Journal (BDJ) and its associated authoring tool, ARPHA Writing Tool (AWT), launched within the ViBRANT EU Framework Seven (FP7) project (Smith et al. 2013). The Biodiversity Data Journal was the first ever journal that provided a fully Web- and XML-based life cycle of a manuscript, starting from authoring to submission, peer review, publishing and dissemination. Later, the BDJ workflow was upgraded to the "ARPHA-XML journal publishing workflow" which itself is a part of the ARPHA Journal Publishing Platform (Penev 2017). The ARPHA-XML workflow came with several tools and workflows developed by Pensoft, such as ReFindit for discovery and import of literature and data references, import/export of tabular data and also of Darwin Core occurrence records, conversion of Ecological Metadata Language (EML) metadata into manuscripts, automated archiving of articles and sub-article elements in Zenodo and others (for details, see next section).

The third stage of Pensoft's effort towards open science publishing was the launch of the Research Ideas and Outcomes (RIO) journal that publishes all outputs of the research cycle, beginning with research ideas; project proposals; data and software management plans; data; methods; workflows; software; and going all the way to project reports; research and review articles, using the most transparent, open and public peer review process (Mietchen et al. 2015). The RIO Journal publishes open science collections of various project or research cycle outcomes, with the EU BON project collection, entitled Building the European Biodiversity Observation Network (EU BON) Project Outputs, being a fine example.

Eventually, all these years spent in development of novel approaches to publication of biodiversity data resulted in a set of standards, guidelines, workflows, tools, journals and services which we define here as ARPHA-BioDiv: A Toolbox for Scholarly Publishing and Dissemination of Biodiversity Data (Fig. 1). The toolbox is designed to ease scholarly publishing of biodiversity and biodiversity-related data with special emphasis on the EU BON and GEO BON networks. ARPHA-BioDiv constitutes a key EU BON deliverable (D.8.3).

Figure 1.

ARPHA-BioDiv is a set of standards, guidelines, tutorials, tools, workflows, journals and services, designed to facilitate the scholarly publication and dissemination of biodiversity data.

The market for online collaborative writing tools has long been dominated by Google Docs. However, as it is too generic, it has not met the specific demands of academic publishing and, in recent years, some start-ups have developed platforms and services to fulfil this increasing gap in the publishing market. Some examples include Overleaf (originally WriteLaTeX), Authorea, ShareLatex and others, most of them being based on LaTeX, but differing in the level of complexity and features for manuscript writing. For people unfamiliar with LaTeX, the learning curve is steep which explains the comparatively restricted usage, mostly centred around the LaTeX community. Currently, none of the above-mentioned tools provides all the components of an end-to-end authoring, peer review and publishing pipeline. For instance, most tools lack a peer review system and rely on integrations with well-established platforms, such as Editorial Manager, ScholarOne, or others.

ARPHA has emerged as the first ever publishing platform to support the full life cycle of a manuscript, from authoring through submission, peer review, publication and dissemination, within a single, fully Web- and XML-based, online collaborative environment. The acronym ARPHA stands for "Authoring, Reviewing, Publishing, Hosting and Archiving" - all in one place, for the first time. The most distinct feature of ARPHA, amongst others, is that it consists of two interconnected but independently functioning journal publishing platforms. Thus, it can provide to journals and publishers either of the two or a combination of both services by enabling a smooth transition from the conventional, document-based workflows to fully XML-based publishing (Fig. 2):

Figure 2.

ARPHA consists of two independent journal publishing workflows: (1) ARPHA-XML, where the manuscript is written and processed via ARPHA Writing Tool and (2) ARPHA-DOC, where the manuscript is submitted and processed as document file(s).

1. ARPHA-XML: Entirely XML- and Web-based, collaborative authoring, peer review and publication workflow;

2. ARPHA-DOC: Document-based submission, peer review and publication workflow.

The two workflows use a one-stop login interface and a common peer-review and editorial manuscript tracking system. The XML-based workflow in use at Biodiversity Data Journal (BDJ) was the first of its kind back in 2013 and has since seen continuous refinement over the course of more than three years of active use by the biodiversity research community. It is also now used by the Research Ideas and Outcomes (RIO), One Ecosystem and BioDiscovery journals. The second, file-based submission workflow, is currently used by ZooKeys, PhytoKeys, MycoKeys, Journal of Hymenoptera Research, Nature Conservation, Deutsche Entomologische Zeitschrift, Zoosystematics and Evolution, NeoBiota and other journals, published by Pensoft.

At the core of the ARPHA-XML workflow is the collaborative online manuscript authoring module called ARPHA Writing Tool (AWT). AWT’s innovative features allow for upfront markup, automisation and structuring of the free-text content during the authoring process, import/download of structured data into/from human-readable text, automated export and dissemination of small data, on-the-fly layout of composite figures and import of literature and data references from online resources. ARPHA-XML is also perhaps the first journal publishing system that allows for submission of complex manuscripts via a dedicated API.

The generic and domain-specific features of ARPHA (used for publication and dissemination of biodiversity data via the ARPHa-BioDiv toolbox) are listed in Table 1 and Table 2 respectively.

Table 1. Download as CSV 

Generic features of the ARPHA Journal Publishing Platform

FEATURE	ARPHA-DOC	ARPHA-XML
ARPHA is a combination of software platform and a wide range of associated services.	X	X
ARPHA serves individual journals or multiple journal platforms.	X	X
Integrated with the industry leading indexing and archiving platform (see list) through web services, APIs and data exchange protocols.	X	X
Individual journal website design.	X	X
Customisable submission module.	X	X
Peer review and editorial management system.	X	X
Peer review process customisable by journal. It can be conventional (either single-blind or double-blind), community-sourced, or public.	X	X
Online collaborative authoring tool (ARPHA Writing Tool, abbreviated AWT, formerly Pensoft Writing Tool, abbreviated PWT), closely integrated with submission, peer review, production and dissemination tools.		X
Collaborative work on a manuscript with co-authors; external contributors, such as mentors; pre-submission reviewers; linguistic and copy editors; or colleagues. The external contributors are not listed as co-authors of the manuscript.		X
Large set of pre-defined, but flexible article templates covering many types of research outcomes.		X
Online search and import of literature or data references; cross-referencing of in-text citations; import of tables; upload of images and multimedia; assembling images for display as composite figures.		X
Automated technical validation step (it can be triggered by authors any time) checks the manuscript for consistency and for compliance with the JATS standard as well as the journal's requirements.		X
Human-based, interactive pre-submission technical check and validation tool helps authors to proceed with their manuscripts to a form almost ready for publication.		X
Pre-submission external peer review(s) performed during the authoring process. The pre-submission peer reviews are submitted together with the manuscript to prompt editorial evaluation and publication.		X
For editor's convenience, peer reviews in ARPHA are automatically consolidated into a single online file that makes the editorial process straightforward, easy and comfortable.		X
In the ARPHA-XML workflow, authors can publish updated versions of their articles anytime.		X
Automated archiving of all published articles in Zenodo and CLOCKSS on the day of publication.	X	X

Table 2. Download as CSV 

Domain-specific features of the ARPHA-BioDiv toolbox used for publication and dissemination of biodiversity data

FEATURE	ARPHA-DOC	ARPHA-XML
Markup and visualisation of all taxon names used in the text.	X	X
Markup and visualisation of taxon treatments following the TaxPub XML schema (an extension of the Journal Archiving Tag Suite (JATS) used by PubMed and PubMedCentral).	X	X
Markup and automated mapping of geo-coordinates of geographical locations.	X	X
Markup and visualisation of biological collection codes against the Global Registry of Biological Repositories (GRBIO) vocabulary (Schindel et al. 2016).	X	X
Pre-publication registration of new taxa in ZooBank, IPNI or Index Fungorum (as relevant).	X	X
Dynamic, real-time creation of online profile for each taxon name mentioned in an article through the Pensoft Taxon Profile tool.	X	X
Automated linking through the Pensoft Taxon Profile tool of each taxon name mentioned in an article to various biodiversity resources (GBIF, Encyclopedia of Life, Biodiversity Heritage Library, the National Center for Biodiversity Information (NCBI), Genbank and Barcode of Life, PubMed, PubMedCentral, Google Scholar, the International Plant Name Index (IPNI), MycoBank, Index Fungorum, ZooBank, PLANTS, Tropicos, Wikispecies, Wikipedia, Species-ID and others).	X	X
Workflow integration with the GBIF Integrated Publishing Toolkit (IPT) for deposition, publication and permanent linking between data and articles of primary biodiversity data (species-by-occurrence records), checklists and their associated metadata.	X	X
Workflow integration with the Dryad Data Repository for deposition, publication and permanent linking between data and articles of datasets other than primary biodiversity data (e.g. ecological observations, environmental data, genome data and other data types).	X	X
Export of XML-based metadata and TaxPub XMLs of the papers to PubMedCentral.	X	X
Automated export of all taxon treatments (new taxa and re-descriptions, including images) to Encyclopedia of Life. Example: http://eol.org/pages/21232877/overview.	X	X
Automated export of all taxon treatments (new taxa and re-descriptions) to Plazi TreatmentBank. Example: http://tb.plazi.org/GgServer/html/B07E9CD77F60DCC65C10A381F6E3BBF0	X	X
Automated export of all taxon treatments (new taxa and re-descriptions), including images, keys, etc. to the Wiki repository Species-ID. Example: http://species-id.net/wiki/Spigelia_genuflexa.	X	X
Automated export of the occurrence data published in BDJ into Darwin Core Archive (DwC-A) format (see also Baker et al. 2014) and its consequent ingestion by GBIF. The DwC-A is freely available for download from each article's webpage that contains occurrence data.		X
Automated export of the taxonomic treatments published in BDJ into Darwin Core Archive. The DwC-A is freely available for download from each article's webpage that contains taxonomic treatments data.		X
Automated export and archiving of images from the published articles in Zenodo. Images from biodiversity journals are imported into the Biodiversity Literature Respository (BLR) of Zenodo.	X	X
Import of Darwin Core-compliant primary biodiversity data from spreadsheet templates or via a manual Darwin Core editor and consequent publication in a structured downloadable format (Smith et al. 2013, Robertson et al. 2014, Wieczorek et al. 2012).		X
Direct online import of Darwin Core-compliant primary biodiversity data from GBIF, Barcode of Life, iDigBio, and PlutoF into manuscripts (Senderov et al. 2016).		X
Multiple import of voucher specimen records associated with a particular Barcode Index Number (BIN) (Ratnasingham and Hebert 2007) from the Barcode of Life.		X
Automated generation of data paper manuscripts from Ecological Metadata Language (EML) metadata files stored at GBIF Integrated Publishing Toolkit (GBIF IPT), DataONE and the Long Term Ecological Research Network (LTER) (Senderov et al. 2016; for details, see also Pensoft's blog).		X
Novel article types in ARPHA Writing Tool: Taxonomic Paper, Data Paper, Software Description, Monitoring Schema, Ecosystem Inventory, Ecosystem Service Inventory, Ecosystem Service Models, Species Conservation Profile, compliant with the IUCN Red List (Cardoso et al. 2016), Alien Species Profile, compliant with the IUCN Global Invasive Species Database (GISD) and others.	X*1	X
Nomenclatural acts modelled and developed in BDJ as different types of taxonomic treatments for plant taxonomy.		X
Automated archiving of all biodiversity articles in the Biodiversity Literature Respository (BLR) of Zenodo.	X	X

Research articles have traditionally been containers for scientiifc results for several centuries and this holds even more for research books. The Internet era brought disruptive changes to academic publishing and one of these is that the notion of the research article as the only valid output for scientific endeavours was challenged. Resulting from this, novel article formats started to proliferate in an attempt to publish extra research objects from across the research cycle, such as methods, data and software. Pensoft pioneered several novel article formats with the launch of the Biodiversity Data Journal. Currently, the ARPHA Writing Tool supports nearly fifty article formats (Fig. 3), used in the Biodiversity Data Journal, Research Ideas and Outcomes, One Ecosystem, and BioDiscovery. The article formats can be generic, e.g. used within almost any domain (for example, research idea, research article, data management plan and others), or domain-specific, such as the article formats described below.

Figure 3.

Article formats available in ARPHA Writing Tool.

Data Paper

Software Description

R Package

Monitoring Schema

Species Conservation Profile (SCP)

Alien Species Profile (ASP)

Ecosystem Inventory

Ecosystem Service Mapping

Ecosystem Service Models

In 2010, ZooKeys published its 50th issue Taxonomy shifts up a gear: New publishing tools to accelerate biodiversity research in a new format based on pre-publication tagging of biodiversity-specific terms in the article XML and semantic enhancements to the published paper (Penev et al. 2010b, Penev et al. 2010a). ZooKeys implemented the TaxPub XML schema, developed by Plazi, later endorsed as an extension of the Journal Archiving Tag Suite (JATS) standard (Catapano 2010). Since then, all life science journals published by Pensoft use the semantic markup workflow in their everyday editorial work to "atomise" and disseminate the content at sub-article level. A list of tools and features for semantic tagging and enhancements of the article content is available in Table 2; implementation and use cases are reviewed by Penev et al. (2012). Examples of the use of the domain-specific markup are illustrated in Fig. 4.

a

b

c

d

Figure 4.

Examples of use of the domain-specific XML markup in the published artices.

a: Interactive mapping of geo-coordinated species occurrences (example from Frolov and Akhmetova 2013).
b: Pensoft Taxon Profile (PTP) is created in real time by clicking on any taxon name mentioned in an article (in this case Annoniaceae from Hoekstra et al. 2016).
c: Images and pages from historic literature where a taxon name has been mentioned are available from various sources (e.g. Encyclopedia of Life and the Biodiversity Heritage Library via Pensoft Taxon Profile (PTP) (in this case Annoniaceae from Hoekstra et al. 2016).
d: All taxon names usages (TNU) in an article are indexed and matched to their type of use (e.g. citations in the text, heading a taxon treatment, associated to images or present in identification keys, example from Brown et al. 2017).

The "integrated narrative and data publishing", or "integrated data publishing", is a relatively new approach, assuming that data or code are imported in a structured form in the manuscript text and are downloadable from the published article. In biodiversity science, this term has been coined and first demonstrated by the Biodiversity Data Journal (BDJ), developed in the course of the EU-funded project ViBRANT (Smith et al. 2013, see also Fig. 5). Publishing of an executable code, also known as "literate programming", in an article was proposed back in 1984 (Knuth 1984), but only recently did we see this practice in journals (Veres and Adolfsson 2011). Another example of integrated data publishing is the linking of a standard article to an external platform that hosts all data associated with the article and provides additional data analysis tools and computing resources; this approach is believed to have been pioneered by the GigaDB and the GigaScience journal (Edmunds et al. 2016). Various kinds of implementing 3D or other multimedia visualisations in an article can also be considered as integrated narrative and data publishing; a good example of that in the biodiversity domain is the paper of Stoev et al. (2013).

Figure 5.

Integrated data and narrative publishing in the ARPHA-XML journal workflow.

Import of Data into Manuscripts

The ARPHA Writing Tool provides online direct import from external databases using community-accepted standards (e.g. within the biodiversity community, these are Darwin Core, TaxPub JATS extension and others - see http://www.tdwg.org/standards/). Initially, data import was from CSV spreadsheets or manually via a Darwin Core HTML editor (Penev et al. 2017). A new functionality of the integrated data publishing system in ARPHA is the online import of specimen records from GBIF, Barcode of Life, iDigBio and PlutoF (Fig. 6). The workflow is described in Senderov et al. (2016). Stepwise guidelines on how to use the feature are also available from Penev et al. (2017) and a blog post.

Figure 6.

Data and metatada import into manuscripts in ARPHA Writing Tool.

Another example of online import of structured text is the ReFindit tool which exists both as a stand-alone application and a plugin in ARPHA Writing Tool. ReFindit locates and imports literature and data references from CrossRef, DataCite, RefBank, Global Names Usage Bank (GNUB) and Mendeley.

Content and Data Export from Published Articles

Article content that is tagged and available in TaxPub XML can be harvested by aggregators which can select and pick sub-article elements, such as metadata, taxon treatments, occurrence records, images and others. Several of these aggregators are major players in biodiversity data preservation and management, for example, GBIF, Encycopedia of Life, Biodiversity Heritage Library, Plazi, Biodiversity Literature Repository at Zenodo, ZooBank, International Plant Names Index, MycoBank, Index Fungorum and many others. The data export in some cases is provided by a featured outbound API. The workflows and aggregators that use the semantically enriched article XMLs are listed in Table 2, and illustrated in part on Fig. 7; the initial core set of features was also reviewed by Penev et al. (2010b) and Penev et al. (2012).

Figure 7.

Extraction and delivery of data and content from published articles to aggregators, nomenclators, archives, and indexers.

All data published in the Biodiversity Data Journal can be downloaded in tabular format (CSV) straight from the article text and re-used by anyone, provided that the original source is cited (Fig. 8). Upon publication, the primary biodiversity data (for example, species occurrence records, species descriptions and taxon checklists) are also automatically exported into machine-readable Darwin Core Archives and become available for harvesting and indexing by aggregators (Fig. 8). Furthermore, species occurrences are indexed and made available as a separate dataset in GBIF bearing the article’s DOI (Fig. 9) which increases the visibility and citation probability of both the article and the underlying data.

Figure 8.

Export of data from articles published in Biodiversity Data Journal. Species occurrences and other structured data tables can be downloaded in CSV format (green arrow); all species occurrences are also available as Darwin Core Archives and are automatically harvested and indexed by GBIF (red box and arrow).

Figure 9.

The occurrence data from articles published in the Biodiversity Data Journal (in this case from the paper of Johnson 2013) are automatically indexed via Darwin Core Archive in the GBIF Integrated Publishing Toolkit.

Data Extraction and Re-publishing Workflow

The present workflow has been created and tested with three different book titles with the support of the EU-funded projects pro-iBiospehere, SCALES and EU BON. It resulted in the launch of the Advanced Books platform of Pensoft, designed to (re-)publish historical or new books in semantically enhanced open access. The workflow is illustrated on the main homepage of Advanced Books at http://ab.pensoft.net and in Fig. 10. Of particular interest was the text and data extraction and conversion to XML of the historical book of Winch (1831) Flora of Northumberland and Durham. Trans. Nat. Hist. Soc. Northumberl., Durham and Newcastle upon Tyne 2: 1-149. Printed by T. and J. Hodgson. The data extraction and conversion has been processed by Quentin Groom from the Botanical Garden Meise, Belgium. The source document was scanned by Ernst Mayr Library of Harvard University for the Biodiversity Heritage Library. The digitised text was uploaded to Wikisource, where it was proofread. The corrected text was then marked-up into XML and semantically enhanced with additional details, including links to the original citations and coordinates of the mentioned localities (for details, see Pensoft blog).

Figure 10.

Data extraction and re-publishing workflow of the Advanced Books platform

A distinct feature of the ARPHA-XML publishing workflow is the possibility to import complex manuscripts, including metadata, text figures, tables, references, citations and others, via an API available in ARPHA Writing Tool (Fig. 11, documentation at http://arpha.pensoft.net/dev/). A working example of the workflow is described in the next section.

Figure 11.

Submission of manuscripts to ARPHA Writing Tool through Application Programming Interface (API).

In order to submit an article via the Pensoft RESTful API, one has first to prepare an XML file according to the Pensoft XML schemas or according to the Ecological Metadata Language (EML) standard (information listed in the link above). An authentication token is obtained from the settings dialogue in the ARPHA-BioDiv which is supplied together with the XML file to the endpoint. If the document is imported successfully, it is created in the respective journal's ARPHA Writing Tool instance, where it can be further edited manually and submitted to the journal.

Data papers, often called also “data articles”, “data notes”, or similar, were first established by the journals Ecological Archives (published by the Ecological Society of America*2) and Earth System Science Data (ESSD) (published by Copernicus) (see Newman and Corke 2009, Chavan and Penev 2011). According to the definition of Chavan and Penev (2011), data papers are “scholarly publications whose primary purpose is to describe data, rather than report a research investigation. As such, data papers contain facts about data, not hypotheses and arguments in support of those hypotheses based on data, as found in a conventional research article. Their purposes are threefold: to provide a citable journal publication that brings scholarly credit to data publishers; to describe the data in a structured human-readable form; and to bring the existence of the data to the attention of the scholarly community.“

The data paper should include several important elements (usually called metadata, or “description of data”), for example:

• Title, authors and abstract;

• Project description;

• Methods of data collection;

• Spatial and temporal ranges and geographical coverage;

• Collectors and owners of the data;

• Data usage rights and licences;

• Software used to create or view the data.

These metadata, if available and deliverable in machine-readable form (XML, JSON, etc.), can be used to produce a “data paper manuscript” that can be submitted to a journal for peer review and publication. The ARPHA approach to data paper publishing was first demonstrated in 2010 in a joint project of the Global Biodiversity Information Facility (GBIF) and Pensoft. As a result, this partnership created a workflow (Fig. 12) between the GBIF’s Integrated Publishing Toolkit (IPT) and Pensoft’s journals (ZooKeys, Phytokeys, Nature Conservation and others). A special module at IPT generates data paper manuscripts into RTF files from the extended metadata descriptions automatically, at the click of a button. Thereafter, manuscripts can be submitted to a journal for peer review and publication. After publication, the data paper’s DOI is linked back to the dataset’s DOI at IPT. In less than three years, more than 100 data papers have been published in Pensoft journals this way (for examples, see the Data paper subsection above).

Figure 12.

Creation of data paper manuscripts from Ecological Metadata Language (EML) metadata hosted at the GBIF IPT

Recently, the workflow was amended by a direct import functionality of EML metadata downloadable from GBIF, LTER and DatONE networks on to a data paper manuscript in ARPHA Writing Tool (Senderov et al. 2016, Penev et al. 2017, see also Fig. 13). The workflow has been thoroughly described in a blog post, while stepwise instructions are available via ARPHA's Tips and tricks guidelines.

Figure 13.

Conversion of Ecological Metadata Language (EML) metadata into data paper manuscripts in ARPHA Writing Tool.

The ARPHA-BioDiv toolbox has been developed in the course of several years and its tools, workflows and journals are used routinely by thousands of authors, reviewers, editors and readers worldwide. It is virtually impossible to list here the numerous use cases and approaches that have been tested and succesfully implemented over the years (see Penev 2017 and Penev et al. 2017 for review). Below we describe three publishing use cases that have been elaborated during the EU BON project.

Expert and Data Mobilisation through the Fauna Europaea Special Issue

Expert and Data Mobilisation through the LifeWatchGreece Special Issue

EU BON Open Science Collection in RIO Journal

Legal Framework and Policies

Licences for publishing and re-use

Strategies and Guidelines for Scholarly Publishing

Tutorials, Manuals and Supporting documentation

In the future, we want to reimagine and reinvent the academic publishing process. At the dawn of academic publishing, papers had been written especially for human consumption. The human mind alone was expected to crunch the data. Now humans rely on computers to store and manipulate the data and verify the correctness of numerical algorithms, whereas our minds focus on the big picture and the story behind the data.

With ARPHA-BioDiv, we have already taken the first few steps in creating articles that can be read both by humans and computers, as has been described so far in this article. However, more can be done. One area of innovation in academic publishing lies in creating linked content - embedding machine-readable database records in each publication that are linked to the world-wide network of linked knowledge hubs. To achieve this goal, we are currently working towards exporting content that has been semantically enriched in a knowledge graph called the Open Biodiversity Knowledge Management System or OpenBioDiv for short (pro-iBiosphere 2014, Senderov and Penev 2016).

This will enable the reader of an aritcle, for example, to connect published occurrence data to portals such as GBIF and geographic repositories such as GeoNames. An illustration of the use-value of this integration will be, for example, an accelerated creation of various models, such as species distribution models, based on the article data. Thanks to the linking of the occurrence data in the article to databases, it will be possible to assemble all the elements needed for a species distribution model of the discussed taxon programmatically in an environment such as R. Moreover, the links in themselves are valuable information and can point to "hot" topics, such as "hot" taxa or "hot" figures, having many incoming links to them (Page 2016). Or, the user may choose to investigate the genetics of the taxon, the occurrence of which they had just seen, through a link to GenBank.

We also believe that a large portion of tradional academic publishing, even if enriched with Linked Data, will be supplemented by nano-publications (Groth et al. 2010, Mons et al. 2011, Chichester 2013). More and more academic research reveals stories and data that cannot be published in the traditional seven-figure-paper. Imagine that the research team you are leading has just discovered 500,000 gene-disease associations across the genome of an important domestic animal. You want all of these findings to be first class research objects - with DOIs, just as publications - and not to be relegated only to a database record that can be altered or deleted. Towards this goal, we are working on nano-publications: first class research objects with DOIs and metadata including author, publisher, etc. which are published as a regular publication, but nevertheless formatted primarily as a machine-readable fact that can be ingested by a database without any alterations.

Finally, we believe that publishers are stewards of the worlds' scientific information and there is knowledge in the totality of the published articles that is not part of any article alone. We are working on artifical intelligence algorithms both from the machine logic domain and from the machine learning domain to discover this hidden knowledge. The authors of tomorrow will have at their disposal not only a tool to format their manuscript, add citations and mark-up their data, but also tools that will discover additional information relevant to the authors' ideas and suggest similar research during the authoring phase. And, if we can dream very big, why not have artificial intelligence algorithms sophisticated enough to act as a research assistant during the authoring phase? What a marvelous thought!

The basic infrastructure for importing specimen records was partially supported by the FP7-funded project EU BON - Building the European Biodiversity Observation Network, grant agreement ENV30845. V. Senderov's PhD is financed through the EU Marie-Sklodovska-Curie Program Grant Agreement Nr. 642241.

LP - vision and management; TG - technical supervision; PG, SD - software development; VS - online import of specimen records and EML metadata; IK, IK - elaboration of tutorials, promotion and PR support; SP - webdesign; PS - editorial supervision and project management.