Revised 5.2.1995.
Most researchers are well aware of the drawbacks of the current scientific publishing process. The time lag between the initial submission of an article and its final publication in printed form is seldom less than one year and can exceed two years. The discussion which follows a paper ("discussion open until ...") may lag the paper itself by half a year, if any such discussion takes place at all. There are also additional obstacles on the route to the final "consumers" of the product. The relatively low number of specialists world-wide who are interested in a narrow research domain leads to a low circulation of the journals in question, which in turn raises the subscription prices. High subscription prices again lead to a lower number of subscribers, creating a vicious circle. The December 1994 issue of Scientific American reports for instance that in the past 8 years journal expenditures nearly doubled while the actual number of journals subscribed to declined by some 10%.
Subscription alone doesn't guarantee that the journals reach their intended readership. Researchers working in large research organisations may receive journal issues only after these have "toured" for months (in some cases even over a year) within the organisation's internal mail system.
Nevertheless the research community needs scientific refereed journals. The information contained in the refereed scientific journals of some domain, seen as a whole, represents the state-of-the-art in that domain better than any other means. Often it takes several years before that knowledge is consolidated and migrates to the basic textbooks of the domain. In areas such as IT in construction basic textbooks of high quality are in fact almost totally missing.
In an era where just about anything can pass as a conference paper the refereed journal paper still represents an original contribution which has undergone some measure of quality control. For many younger scientists getting their work published in refereed journals is extremely important, in particular as a means of building up an academic track record.
All this points to a situation where some fundamental re-engineering of the scientific publishing process is urgently needed. The publishing process needs to be speeded up considerably, especially in research domains where developments are rapid (ie. information technology). Fortunately recent developments in network technology and infrastructure provide tools which are essential for such a reengineering effort.
Commercial publishers of scientific journals have started to look at how they could use IT to their advantage. They are however burdened by commercial considerations which may unduly restrict the solutions they come up with. For this reason they are currently starting to market solutions such as including the contents of numerous journals on CD-ROM:s or offering on-line services with advanced search capabilities including abstracts of articles. In one case the abstracts are put into the service before the article has even been published. They are, however, understandably reluctant to make the articles themselves available in services such as World Wide Web, due to the difficulties in collecting revenues for such a service.
The project plan presented in this paper is the outcome of a small reengineering exercise. The major questions asked were how we could, given the current state of electronic publishing technology;
In the ideal case of a submitted article needing no correction the time from submission to the finished article becoming available could be as short as one month. Also in the case of submissions needing editing the overall process could be shortened considerably. The technical editing time (putting an article on the Web) could be shortened to a few days, compared with the several months required for type setting, proof correction and waiting in line for the first journal issue with empty space.
The journal could in fact provide a virtual meeting place of professionals (readers and writers) interested in the topics of the proposed journal.
At least the following scientific journals are specialised in information technology in construction.
The circulation of these journals varies. The ASCE journal has a subscription base of some 1500-2000 and prints over 3000 copies. Computer-aided design has a large reader base, because of its multi-disciplinary nature. Automation in construction was originally targeted for at least 500 subscribers, to break even. The number of subscribers to some of the other journals is as low as a few hundreds.
A number of periodic international conferences publish proceedings dealing with IT in construction, for instance CIB W-78 annual workshop, CAAD Futures, ISARC, the ASCE congress on construction computing. Often the editors of the above refereed journals solicit papers very actively from researchers who present papers at these conferences.
The quality of the articles published in the above journals varies a lot. One practice which is somewhat questionable is for a journal to publish conference papers almost unchanged. The details of the review procedures and the statistics of the journal (proportion of accepted papers/submitted papers, mean time from submission to printing) are often not known even by the members of the editorial board. To a large extent the responsibility lies with the main editor of a journal.
Electronic publishing uses electronic media to distribute information that would otherwise be printed on paper. "Publications"
differ in modes of distribution and presentation formats.
Just-in-time distribution provides information only on request - when searched for and requested by the end user. Examples of this kind of publications include existing electronic journals, full text bibliographies, databases, on-line conference proceedings and other on-line publishing efforts like Project Gutenberg, Britannica-on line. Computer networks and applications like FTP, Gopher or WWW are used for the physical data transfer. Since the bandwidth for the just-in-time distribution is limited, only timely or not very voluminous information is currently distributed in that way.
Just-in-case distribution generally uses the same channels as traditional publications (mail) but instead of paper employs digital media to store the information. This information is shelved and will be used if and when required. The most popular media for full-text publications are CD-ROM:s which can hold about the same amount of text as 200 Bibles. In particular reference information (Encyclopaedias), databases like (ICONDA or CD-REEF) and full text excerpts from selected computer magazines are published in this way - some on a subscription basis.
Finally once-in-time distribution broadcasts information that is lost soon after it was made available. Examples are non archived e-mail exploders, US Net News etc. Information that should be distributed in that way are very timely messages and pointers to new resources available through the previous two channels.
Electronic publications take many different formats, from simple ASCII text to multimedia including HiFi sound and full motion video. Formats may be proprietary - that is information in that format may only be accessed using particular commercial software. In that case the information is usually tightly bound with the presenting application. Alternatively the format can be in some way standardised so that the information may be extracted by many different, perhaps even free software tools.
Existing electronic journals distributed on the WWW include:
A comprehesive list is maintained at the World-Wide Web Virtual Library: Section on Electronic Journals . A recent article in Science reports over 70 fully electronic, peer reviewed scholarly journals. Scientific American again reports 440 scientific journals and newsletter, 100 of which are peer reviewed.
The proposed journal should make the bulk of its information available on just-in-time basis. It should, however, regularly announce the availability of the information using the broadcast channels. It should be presented using standard formats which can be used on a wide range of platforms over long periods of time.
From the viewpoint of potential authors the following functional requirements on an electronic journal seem reasonable:
From the viewpoint of potential readers the following functional requirements seem reasonable:
It is important that an electronic journal has at least the same or possibly better reproduction capabilities as current printed journals. The following features should be mandatory for both on-screen viewing and print-outs.
The following features would also be useful but are not as essential in the short run to get the journal running. Such features could be added later on.
Mathematical formulas and notations could be handled as vector-based figures. The last two items could possibly be handled in the same way as appendices, by inserting hyperlinks to separate documents on the Web.
The scope of the proposed journal is closely related to the scope of CIB working group W78 (Information Technology in Construction).
The aim is to report on academic research and innovative industrial development work concerning the application of information technology throughout the life-cycle of the design, construction and occupancy of buildings and related facilities. The focus is on presenting research results dealing with or having relevance across disciplines and/or life-cycle stages. There are other, existing journals addressing specialised analysis, synthesis or optimisation methods particular to design or construction tasks within the individual subdisciples of construction.
In the following some major topics which the journal aims to cover are mentioned:
Papers eligible for publication should present original research or be state-of the art reviews. Papers presenting research in the early stages or research ideas only should be avoided. (We should in fact stop talking about "papers" and use articles or information, since paper is the medium we want to avoid as far as possible.)
In addition to university researchers, the journal invites articles from industrial researchers and developers, especially those presenting how the new capabilities offered by IT are being incorporated into software products, or how companies are applying state-of-the-art IT solutions to design and construction practice.
The term electronic (or digital, Internet, on-line ) should be included in the name to emphasise how this journal differs from already existing ones. The suggested full name is Electronic Journal of Information Technology in Construction.
In many connections it is handy to have an abbreviation. Such an abbreviation could also be used as a LOGO in the home page on WWW. In the following some alternatives are discussed:
The following overall organisation is suggested for running the journal.
It is important to have a background organisation with sufficient legal status and an administrative apparatus to manage possible money transactions etc. In the case of paper based journals at least three options are commonly used ( commercial publishers, universities and national professional associations).
The suggestion is that the Royal Institute of Technology could function as official publisher.
Obviously this should be CIB W78. The W78 board will be asked to endorse the journal and W78 members will be filling the main functions of the journal' s organisation. On the cover page of the journal this endorsement should be clearly stated and hyperlinks to the W78 home page should be inserted (and vice versa in the W78 home page). The editors will regularly report to W78 on the progress of the journal during meetings or using email.
This is where the articles, home pages etc. reside physically. Ideally there should be two web sites, one of them being the primary site, and the one which manages all the traffic of submissions etc.
There are two alternatives for working with a such system of two sites. A mirror site implies that part of the readers will use the mirror, and part the primary site. The advantage is that the mirror also reduces network traffic and may speed up access. The disadvantage is that the mirror cannot be identical to the original since the inbound information related to the review process must be centralised.
A backup site implies that the site only pops into existence if the original site is down or can not be accessed. In that case it takes over full operations of the system. The advantage is above all better security, and it is much easier to implement than a mirror site. The disadvantage is that no traffic optimisation takes place.
The suggestion is to start with one or even several backup sites. These can later evolve into mirrors if this is needed due to traffic volume or persistent network problems.
The suggestion is that the primary site should be at the University of Ljubljana. A decision about a secondary site will be made later.
An editor
The editor takes overall responsibility for the journal (in the legal sense). He also has the major responsibility for the review process and for the scientific quality of the journal. He functions as the chairman of the executive board formed by himself and the three associate editors.
The suggested editor is:
Prof. Bo-Christer Björk Royal Institute of Technology, Stockholm, Sweden
Three associate editors
Together with the editor these form the executive board of the journal. All major decisions should preferably be worked out by these four in consensus. One or two of the associate editors take responsibility for the maintenance of the WWW sites and the email traffic. All the associate editors help the editor in managing the reviews of individual submitted papers.
Suggested associate editors are:
Prof. Charles Eastman UCLA, Los Angeles, USA
Asst. Prof. Ziga Turk University of Ljubljana, Slovenia
Dr. Dana Vanier IRC, Ottawa, CANADA
A review committee
The main tasks of the review committee are to:
The review committee provides the core of experts for reviewing the papers. As far as possible the reviewing assignments should be kept within this board. One of the reasons for this is that this may help in achieving consistent quality standards in the review work. Another reason is that the practical arrangements of managing the review process are easier with referees who are familiar with the technicalities for downloading submitted articles etc.
If reviews are predominantly solicited within the committee it would consequently need to be rather large in number. The list should in no way be closed but new members could be added continuously as new referees are recruited.
In the choice of members the following criteria will be used.
The four editors will together compile a list of candidates and ask the included experts for their consent. This list will be published in conjunction with the first call for papers or shortly afterwards.
The submission and publication procedure consists of the following steps:
Articles should be submitted in electronic format only. Submissions on paper or on floppy disks will be rejected and the authors asked to kindly resubmit the article using the correct techniques. Articles should be submitted to the executive board by email or (prefearably) FTP with an accompanying e-mail message. Given the state of technology and the availability of tools "rich text format" seems most appropriate at this moment. Detailed guidelines for the submission procedures and formats will be issued later.
The following requirements should be made on the papers in terms of size, referencing etc.
The editorial board has no facilities for making thorough checks of the language of submitted papers. It is assumed that authors have run the articles through spelling checkers. Foreign language authors are in addition recommended to use native English speakers for checking the language before the submission. A good policy would be to always include at least one native English speaker among the referees.
Formating guidelines and accompanying style sheets for Word for Windows and Word for Macintosh will be made available. The suggested styles will enable uniform look of the document on both the printed (both US Letter and DIN A4 paper sizes will be supported) and WWW versions. The system of heading numbering will also be specified. An example article (taking some already published paper by one of the editors) will be formatted according to the guidelines and put on the WWW for authors
to look at.
The main editor and the three associate editors take turns in being responsible for managing a submitted paper through the refereeing process. The responsible editor assigns three reviewers. If he so wishes he can be one of the three reviewers himself. He collects the answers and makes the final decision concerning the paper. The internet system will be set up in such a way that only the editor and the associate editors will have the authorisation to set up papers in the journal.
The referees can suggest:
The suggested approach to the review process could be based on email communication between the editor and the reviewers. This mode could function is such a way that the three reviewers first work independently without knowing each other's identity. Reviewers will get electronic copies of the submitted papers in exactly the same format as the final published papers. Ideally they should be able to write comments digitally into the papers themselves, but this may be too demanding . Many reviewers will wish to print out the papers, take the stuff home over the weekend and scribble on the printed copy. The suggestion is that the evaluation page is available as an email template.
A question to debate is how much freedom of choice the editor in charge should be given. At one extreme is a situation where the opinions of the referees are only indicative and where the responsible editor has the final say. In a more formalised approach a truth table could be made of the choices open for the editor.
If all three referees indicate acceptance of the paper with or without minor corrections it should be accepted. The responsible editor checks that the typographical corrections, if they have been requested, have been made. Similarly if all three referees suggest either rewriting or rejection that suggestion should be followed.
If the referees differ in opinion the responsible editor should start a short email conference with the three referees, to see if a consensus could be reached (after they have read each others comments on the submitted article). If no consensus is reached in a reasonable time the responsible editor makes the decision based on his own judgement.
Also in the case of an unanimous decision the referees should be allowed to see each others comments ( which is much better than the practice of many current journals, where you get very little feedback on your referee work).
KTH takes care of the ISSN number for the journal. The journal has yearly volumes, which follow calendar years. Within a volume articles are numbered sequentially from 1 to n. There are no separate issues within a volume.
Each paper is put on the web by the responsible editor. Alternatively this is handled by somebody at the web site, after he has gotten the OK from the editor in charge. The full information of an article includes:
The full text for on-screen reading should use WWW's native HTML language. The format should be ergonomic for quick browsing. In addition to straight off reading it should contain hyperlinks to other information available on the web (such as earlier articles published in this journal). In the first issues these hyperlinks should be limited to the list of references section only ( at the end of the paper). The author would supply the links. In addition a hyperlink should be inserted at the end of the paper if any discussion has followed.
Any other more elaborate hypertext schemes are, of course, exciting, but will probably lead to technical problems and too much work for the editors. Such schemes can be added later on if they appear feasible.
The second format is for printing out on a local laser printer. The printouts should look almost the same as today's paper based articles. Postscript has been suggested as the print format.
Authors should be encouraged to place large appendices to the papers as separate items on some WWW site and indicate the route to them in the original paper. This implies that even software demos or short videofilms could be appendices. There are however some technical problems and problems of how strictly such appendices should be reviewed which need to be sorted out.
The journal could contain some add-on material of interest, which would be affiliated with the journal or made very obvious by hyperlinks for the occasional reader.
Especially the discussion part could be quite exciting compared to traditional paper based journals. The discussion should be relatively free with only limited censorship and delays. The party submitting a discussion item should email it to the editor responsible, who should check that the proposed items fulfils some minimum requirements on language, format and courtesy. We should welcome discussions with two-way communication between the commentator and the author of the original articple, but not quite as personalised as some email conference discussions have turned out to be. Discussion items will be advertised in the original paper by a hyperlink (put there later on by the responsible editor) indicating follow-up discussion.
The inclusion of shorter papers presenting research in progress has also been suggested. In the initial stages of the journal authors of such papers should be encouraged to submit these to coming-up conferences such as the W78 conferences, rather than including these in the journal proper. In the near future also w78 proceedings may be made available immediately after the conferences on WWW. In that case it would be easy to have cross referenced hyperlinks between the journal and the proceedings.
Later on, if there are strong indications that readers of the journal wish to have a separate work in progress section, this issue could be reconsidered.
The basic policy is that the journal is free of charge. Subscription could be in the form of enlisting on a particular email list which regularily disseminates messages on what articles have been published. Alternatively existing email channels could be used for these messages (which may lead to people getting the same mail several times). At least in the first year it might be a good idea using existing channels to get people aware of the journal.
The question of copyright is complicate. Normally a commercial journal secures the right to publish the article in other formats, such as anthologies etc. Special permission has to be asked to be asked for any subsequent publications. In this context a looser policy seem appropriate, say that the author has free right to republish the article in other media, provided that it is clearly stated in any subsequent publication where the article was first published. He would thus for instance have the right to include it in his university's local WWW server if he so wishes.
The journal would on the other hand have the option of later publishing bound paper volumes of the journal, CD-ROMs etc. without getting additional consent from the author.
A rather good formulation which could be used as a basis for defining a copyrigth clause is the Architronic one:
Authors of contributions to Architronic retain copyright and all intellectual property rights appertaining to their works, including but not limited to all rights to authorise subsequent publication and/or translation. Any factual inaccuracies or opinions expressed therein are the authors' own, and do not necessarily reflect the knowledge, views, or positions of Kent State University, any of the university's units, or Architronic's editors.
Material appearing in Architronic may be distributed freely by electronic or any other means, providing that any such distribution is without charge (unless for purposes of cost recovery by interlibrary loan services) and that Architronic is acknowledged as the source. However, no article may be reprinted in any publication without the explicit written permission of the author(s). This statement must accompany all distributions of Architronic, whether complete or partial.
The question of publishing in a paper format is tricky. One possibility is to start off with annual or semi-annual bound volumes (the periodicity should be determined so that the thickness and weight of the separate volumes make them easy to handle). These could be produced as cheaply as possible using offset. The idea would be to sell these at a price which is comparable to the prices asked by commercial refereed papers for a years subscription. The marketing should be handled as cheaply as possible.
The problem is that if we start parallel paper-based printing and find that it's not economical or otherwise feasible it's difficult to stop after a short time. Another option is to start only electronically and make printed copies of the journal available if there is sufficient interest. What will be provided in any way for research groups and libraries wishing to make paper copies is the volume title page which could be placed on top of a binder, as well as the table of contents of the volume in question.
In any case archival laser printed originals of all papers should be stored in safe places such as the KTH library. In real emergencies the library would be allowed to sell paper copies of articles according to the normal way a technical library functions.
The affiliation with CIB W78 is fairly clear. On a political level it could be a good idea to inform the CIB central secretariat in Rotterdam about the journal before the journal starts operating. From CIB's viewpoint the journal could be an interesting pilot and other working commissions may want to follow suit.
The marketing of the journal should, as a rule, be done electronically only. Some exceptions could be made to this rule in the initial stages. A press release should be made and sent to organisations such as CICA, ACADS, ACADIA etc. for publication if they so wish, especially to attract attention from industry. The journal could also be presented in a flyer to be distributed at some major conferences this year (ASCE, W78, CAAD futures, etc.).
In order to keep an open policy the journal should regularly publish statistics on key facts such as: