Peter Fisher's Valediction is very interesting and promising for those who work in the area of geospatial data modelling (or representation) as It is the
foundation of all else that is possible or can be done.
A pdf version of Fisher's Valediction can be accesses from
http://www.informaworld.com/smpp/content~content=a784379045~db=all~order=page or you can read the html version below:
IJGIS Valediction1. Introduction
I have been editing the International Journal of Geographical Information Science for the last 14 years. I was first associated with the journal in this role in 1994 during the publication of volume 8. Then, it was publishing 6 issues per year with a target length of 600 pages, which allowed approximately 30 articles to be included. In 1996, this increased to 800 pages, and in 2005 to 1200 pages. The journal now carries nearly 60 articles per year, and will have a modest increase in volume again next year.
I took over the journal from the capable hands of Professor Terry Coppock, who worked diligently to establish the journal as one which would persist and become the premier journal of record for those working on the development and application of geographical information systems, whatever their background. I have endeavoured to maintain the status of the journal, and I believe that I have. Elsewhere (Fisher 2007b), I have listed some of the competitor journals of IJGIS. IJGIS is unusual because it is listed in the Journal Citation Rankings of ISI's Web of Knowledge in four subject areas (Geography, Physical Geography, Information Science, and Computer Science). Among competing journals, IJGIS has had the highest impact factor over a number of years.
Volume 11 saw the journal published under its present title, when the name was changed from IJGISystems to IJGIScience, in recognition of the fact that the journal had always been engaged in the publication of research into the science of geographical information which underpins the systems that are in widespread use.
2. A personal view of research published in IJGIS
The research that has been published in IJGIS over the years can be divided many ways, but I choose to look at it as is illustrated in figure 1. These are the themes I see which have persisted through the 14 years. The structure identified here was first articulated at a presentation at the AGILE 2007 Annual conference in Aalborg. I would like to thank the organizers (including Lars Bodum and Monica Wachowicz) for inviting me to give that presentation.
Figure 1. Personal view of general research topics published within IJGIS.
To me, the most important research theme is that of Representation. It is the foundation of all else that is possible or can be done with geographical information. I view it as having five components:
Spatial Information Theory addresses how we conceptualize spatial information, and is absolutely central to GIScience. It has been a persistent theme with issues of RESELS, geoatoms, object orientation, and multiscale and multiresolution information as part of it.
Issues of Uncertainty in its broadest sense may be the most common research topic published in IJGIS. This includes probabilistic and fuzzy formalisms, error modelling, rough sets, and semantic uncertainty, among others.
Researchers have long bemoaned the lack of Temporality in geographical databases, but over the 14 years, many papers have been published in this area.
IJGIS has not been slow in publishing the results of Ontological research both from a database construction point of view and from a semantic understanding point of view.
Finally, and perhaps a smaller component than is desirable, is the research on Geometric representation.
The second broad topic is modelling, which, for convenience, I divide into:
analytical and statistical modelling, including network modelling and spatial statistics; and
process modelling, including modelling of social and environmental processes and the technology of those models.
Visualization has always been a major theme within GIScience, and of course, cartography, and computer cartography in particular, is one of the antecedents of the field. Many interesting papers and special issues have been published on topics from this field, including generalization, visual analytics, geocollaboration, and interactive mapping.
Cognitive studies and usability are concerned with how we relate to the world and the information about that world. There are increasing studies on usability, but my personal view is that studies on spatial cognition should ground much research in GIScience but have not been published in IJGIS, with a very few exceptions. I hope that the future may see more such research linking these areas.
A final persistent theme has been that of data policy with which I bracket social construction of information. The first has been researched in many ways and most recently within the umbrella of Spatial Data Infrastructures. The social construction argument is seen by some as anti-scientific, but in my view it is part of all information, as some recent studies have demonstrated, and those studies have shown some potential for working with different world views within GIScience.
Over the 14 years, paper submissions on some research topics have ceased. Parallel processing on which a special issue was published in Volume 10 has become a low-level system issue, with barely a mention of the topic in more recent issues of IJGIS. Similarly, Interoperability was the topic of a special issue in Volume 12, but it too has not been addressed directly in much writing in IJGIS since. The topic remains important, but within the research published within IJGIS, it has been subsumed within the interoperability of data, or within the developing area of Web technologies. Another person might see papers on Web technologies as another emerging component of the IJGIS research literature, but currently I see the Web issues as one that touches many of the other topics raised, particularly Visualistation and Data Policy.
Discussion of the structure outlined here has led others to suggest to me that the World Wide Web, Location Based Services and Global Change are so called 'killer applications' for GIScience, and so might be viewed as themes for structuring the field. These are all interesting areas for research with their own challenges and problems. However, I would rather see these as important areas for application, along with many others, rather than as driving forces. I believe that when an application becomes a driver, it moulds the science, and I do not believe that all applications will fit one mould. Therefore, it is necessary to keep the independence of the core issues as central to GIScience, and not view GIScience as issues of any one application.
3. Issues in producing the Journal
Many issues could be mentioned in the production process, but two stand out for me. The first is with respect to reviewers, and the second is the preparation of graphics.
3.1 Problems with reviewers
The most intractable problem in managing any peer-reviewed journal is making timely decisions on articles. This process is a trade-off between the need for reviewers to have time to read a paper, and an author's wish to have a rapid response, as well as the editor's wish to 'have a life' and do some of their own research. The most frustrating part of managing the process is that reviewers repeatedly promise to complete a review within a particular time period, but fail to do so. This can be for understandable reasons, but when the reviewer then promises to do the review by some new date but fails, and promises again and again, the process becomes very frustrating for everyone.
When I first started editing, a member of the editorial board said to me 'I hope to complete three reviews for every paper I publish'. 3 to 1 is the minimum ratio of reviewed to published papers to which all active researchers should commit. Because of rejections, unfortunately the ratio actually needs to be considerably higher. Unfortunately, there are people who will never return a review, no matter how many times they promise, and there are others who will always return a review, once they have said they will. Research productivity and administrative responsibility is no indicator as to group membership—some of the busiest people are the most reliable. But if you publish one paper, you should commit to reviewing at least three papers, and you should do them as if you were the author, in a prompt and timely manner.
3.2 Problems in graphcs
Authors should be more careful in their design of graphics, graphs, and maps. Perhaps the worst are the graphs generated in modern spreadsheets. One particular spreadsheet package uses grey backgrounds so that graphs are highly visible on the screen, but when these are printed, the grey tends to obscure the actual graph, as do such ephemera as the grid lines and oversized point markers. Unfortunately, many authors seem to be ignorant of the design guidelines of Edward Tufte (1983), which should be studied with care by all involved in illustrating scientific articles. Authors should be prepared to make multiple changes to graphs in the process of preparing an article, using smaller symbols and clear, white backgrounds. Similarly, many authors use grey fills for boxes in flow diagrams. On the whole, these are completely redundant and only obscure the text within the boxes. Boxes should be white, with the outline used to code the boxes, if that is desirable.
In the print technology used by the publishers, colour continues to be expensive, but colour in the electronic version of papers is free. This means that as much colour as an author wishes to include can be carried in any article, but the print version of that article may include all those graphics in greys. The problem with this is that many colours will produce the same grey, so that if information is colour-coded, but the print version is in grey, the coding may not carry over. Therefore, authors need to continue to be be careful in their use of colour and, where necessary, may need to consult experts in the use of colour.
4. Is it still research?
There are many interesting and challenging research topics to be addressed in GIScience, but there are some topics which might be considered to be pass for publication in IJGIS. Without wishing to put off researchers, I would like to mention two here.
First is the annual assault on the editors of papers documenting yet another instance of a raster-GIS implementation of the Universal Soil Loss Equation (USLE), by authors who have not read the literature well enough. This topic was first addressed in the 1980s by, for example, Spanner et al. (1983), and papers being submitted in 2007 are very little different. I am not saying that soil erosion modelling is pass, but as scientific research the USLE is, both within and outside GIS. On the contrary, research relating to more advanced soil erosion models is welcome, and excellent contributions have been included in the Journal, when they meet the review standards.
Similarly, many articles have been written on comparisons of a modest number of surface interpolation algorithms in an experimental situation (whether from point observations or contours and using IDW, spline, and kriging, perhaps). Papers continue to be submitted doing no more. It is easy to conceive of such an experiment, but it is a real challenge to make it original and different from previous experiments, and to demonstrate that the conclusions can be generalized to other contexts. Generation of digital elevation models is no longer dependent on the interpolation of values from sparse point observations or contour lines, but has moved over to measurement-based remote sensing devices such as Lidar and Ifsar. Interpolation remains important for these technologies, but the issues have changed. Future experiments need to be demonstrably relevant.
5. Thanks
During the 14 years I have been working on the journal, I estimate that about 625 papers will have been published, which means that something of the order of 1800 papers have been submitted. A number of people in various roles have been involved, and I would like to record my thanks to them all (in spite of having listed many in a previous acknowledgement; Fisher 2007a):
First are those people whose work has been published in the journal over the last 14 years. I thank them for taking the time to conduct the interesting research they have submitted, and for writing it up. Almost without exception, they have taken criticism from reviewers and papers have gone through changes in the review process. We believe that the published papers which result are better than those originally submitted, but making the changes can be nonetheless painful for the authors. It has been a pleasure for my colleagues and I to see this work through the review process.
Because each paper is sent to at least three reviewers, approximately 5400 requests for reviews have been dispatched. I am ashamed to say that I have no idea how many reviewers this equates to, because I do not know how many have been asked more than once, although I suspect it is the majority. My thanks go to all those who have responded with reviews, when requested. The work involved in taking time and care to consider and critique the work of others cannot be understated, but it can also be most rewarding. Foremost among these reviewers have been members of the editorial board.
All journals have two classes of author: authors whose work is accepted, and those whose work is rejected. The acceptance rate is approximately 30% of submissions, and therefore the latter group is about twice the size of the former (except that, of course, some authors are in both categories), and having taken the effort to conduct the research and write the paper to then have it rejected for publication is always very dispiriting. These are the unacknowledged facilitators of the peer-review process, and I would like to take this opportunity to thank them all, because their work has come to nothing and will not be published in this journal.
During the 14 years, 16 special issues have been published, and a number more are in preparation. The editors of these issues are numerous, but they are acknowledged by being the authors of guest editorials.
It has been my pleasure to work with a number of other people in editorial roles, including Eric Anderson, Steven Guptill, Marc Armstrong, Harvey Miller and now Mark Gahegan as North American Editors (now Editor for the Americas), and Dave Abel and now Brian Lees as editors for the Western Pacific (now Editor for Australasia and Eastern Asia). I have worked with Neil Stuart, Nick Tate, and Lex Comber as Book Review Editors.
Throughout my period as editor, the Publisher's principal representative has been Richard Steele. Direct managerial contacts for the journal have been Meloney Bartlett, Rachel Sangster, and Virginia Klaessen. On the production side, managing the work of anonymous typesetters and copy editors, are the people with whom authors have communicated about proofs (whether they know it or not). They have been David Chapman, Sophie Middleton, Heidi Cormode, and currently James Baldock.
Finally, I must thank Jill Fisher, who has given continuing support and assistance in communicating with authors and reviewers.
The system of peer review, which is the current paradigm for scholarly publication, would not work without all these players; all are crucial to the process. My thanks to all these people in their various roles, from reviewers and authors, to editors and production managers, and to anyone else I should have named but have not. The last 14 years would not have been possible without each and every one of you.
I would like to close by offering my very best wishes for continuing success of the journal to the future editorial team, including Brian Lees (Australian Defense Force Academy, University of New South Wales) as both Editor in chief and Editor for Australasia and Asia, Mark Gahegan (Pennsylvania State University) as Editor for the Americas, and Sytze de Bruin and Monica Wachowicz (Wagenigen University) as Editors for Europe and Africa. I hope that they find it in as good condition as Terry Coppock left it for me.
References
1. Fisher, P. F. Fisher, P. (ed) (2007a) Preface.. Classics from IJGIS: Twenty Years of the International Journal of Geographical Information Science and Systems pp. v-vi. Taylor & Francis , London
2. Fisher, P. F. Fisher, P. (ed) (2007b) 20 years of IJGIS: Choosing the classics.. Classics from IJGIS: Twenty Years of the International Journal of Geographical Information Science and Systems pp. 1-6. Taylor & Francis , London
3. Spanner, M. A., Strahler, A. H. and Estes, J. E. (1983) Soil loss prediction in a geographic information system format.. — In Papers Selected for Presentation at the 17th International Symposium on Remote Sensing of Environment. Volume 1, pp. 89-102. (Environment Research Institute of Michigan, Ann Arbor.)
4. Tufte, E. R. (1983) The Visual Display of Quantitative Information Graphic Press , Cheshire, CT
Posts
