Ontogenesis: One Year On

Abstract

The Ontogenesis kblog has now been available for a little over a year. This offers us a good time to sit back and look upon what has been achieved, and how we can improve things for the future.

For the reader

The original motivation for Ontogenesis came from our long-held desire for a book on ontology development, aimed at a level appropriate for incomers into the field. As a discipline, we have made some great strides in the uptake of ontological technology, but to those coming afresh to the field it is often difficult to know where to start.

However, book publishing tends to be a long-winded and difficult process; we wanted something lighter, faster, responsive and easier. Our experiences with blogging technology suggested this might be a good way forward. At the same time, we wanted Ontogenesis to have an appropriate degree of academic credibility; we didn’t want a collection of opinion pieces or unformed ramblings that define much of the blogosphere. We wanted to see if we could use a blog engine to replace a process akin to the existing publishing process.

The Ontogenesis kblog now offers a collection of small accessible articles on aspects of knowledge and semantics, primarily ontologies, within biology/bioinformatics. At the end of the first year we have over twenty articles available in the Ontogenesis k-blog. There is an mean “reads per month” of 1,000, which makes for a total number of accesses of over 14k, with a peak of 1,250 reads for Jan 2011. The most accessed article, perhaps unsurprisingly, is What is an Ontology.

Even given that this is after robot removal, this is clearly not a statistic to set the publishing world alight; however, it is of a similar size to the number of page reads that might be expected for an “average” PLoS one article, and it is also reasonable when compared to the print run for an average academic book. In short, we feel that these statistics are rather good.

For the reviewers

Whilst content supply has been fairly good, the gathering of reviews by authors has not been so good. In the Ontogenesis k-blog, authors are supposed to manage the review process themselves. Most authors, including ourselves, just stop after publishing an article. this may be for several reasons:

The first and we think the most important reason relates to an issue identified early by Sean Bechhofer: are we a blog or a wiki? In this web-based environment, authors are more used to a collaborative style of writing. For instance, when I (PL) started an article which involved a description of the semantics of OWL; as I am not an expert, I asked Uli Sattler to be a co-author, partly to check my initial statements. This has improved the quality of the article but, of course, she was now no longer suitable as a reviewer.

A second, more social reason, is that authors like to write and publish; once that is achieved the drive to gain reviews is lost. If publication on the kblog didn’t happen until there were reviews, the incentive would change. This would, however, interfere with the lightness of the process. Of course, authors want comment on their articles and kblogs, like any blog, have such a mechanism via the standard commenting mechanisms. We will also have to look at incentives for getting authors to gather reviews; even simply emailing people to “come, look and comment”.

A third, and critical reason, is that the system we had for peer-review was hard to use. We replicated a complete “two reviewers and editor” style system. Authors, reviewers and editors had to remember where they were. This was just too difficult.

For Ontogenesis, we have addressed this in two ways; we have moved to a single, author directed reviewing system. Authors choose their own reviewer, and decide when they have addressed the comments; as all the information is public, we consider that this should be sufficient. Secondly, we have provided some software support within WordPress to help manage the process.

There are many forms of reviewing besides the standard peer-review; Wikipedia has system, while H2G2 has another. Perhaps, in the end, we need to think less of authors and reviewers, and more of primary and secondary authoring. It is not clear to us what the future holds for the peer-review system in science, but we hope that the kblog as a platform is a suitable place to experiment.

For the authors

Our original intention was to provide a resource that was equivalent to a book. However, based around blogging software as it is, Ontogenesis allows us to take advantage of a rapid publication framework; this, in turn, has changed the form of articles that we are writing. While, we still have “long-form” articles such on topics such as health informatics, as authors we have started to write shorter articles on small-discrete topics such as closure axioms. This form of article is, we hope, useful; that James Malone and Helen Parkinson’s article distinguishing reference and application ontologies has had 500+ page reads suggests that they are. They are also enjoyable to write.

We have also seen the advantages over a wiki style environment. Authors can provide their own ideas within an article, as well as adopt their own style. Although it is a small advantage, that an article can contain mild humour, in the end, produces a more readable resource. Our experience with Wikipedia suggests that this form of personal approach is rapidly edited away. Of course, the brutal reality is the main advantage is that the work remains attached to the individual authors; whether or not we like it, academics need to self-promote and cannot afford to do work without credit.

For the future

At the start of the Ontogenesis k-blog, we used a vanilla WordPress installation. This worked, but needed some extras to make it suitable for scientific publishing. We have been fortunate enough to recieve funding from JISC for kblog. Work on this grant is now well underway; we now improved how-to documentation, we have a more refined process, and several pieces of software for presentation of maths, for citations and to enable searching, sharing and storing. Ontogenesis is now archived by the British Library, articles have DOIs, and we have been included in Google Scholar. We are hoping for inclusion in PubMed. In short, we are filling the gaps between Ontogenesis and a formal academic book rapidly, without compromising on the original vision of a rapid, and easy-to-use publication framework.

We feel that this form of publication has a strong future ahead of it. The experience of getting this content onto the web has been enjoyable, and informative. We hope that the resource is valuable to others and that authors continue to find a desire to contribute.