Next: The EUVE Knowledge Base
Previous: The XTE FITS Database (XFDB)
Table of Contents --- Search ---
PS reprint
Peter B. Boyce
American Astronomical Society, 2000 Florida Ave., #400,
Washington, DC 20009
Chris Biemesderfer
ferberts associates, 1805 Mt. Lemmon Rd., P. O. Box 1630, Oracle,
AZ 85623-1630
The AAS was founded to promote communication of research techniques and results within the astronomical community. While this mission has been enlarged through the years, the fostering of effective communication through the publication of high quality scholarly research journals and the organization of scientific meetings remains the prime purpose for the existence of the Society. The major AAS journals, the Astrophysical Journal together with the Astronomical Journal, comprise over half of the world's peer reviewed astronomical literature.
The term ``Electronic Publishing'' means different things to different people. Authors tend not to consider the parts which come after the preparation of the original manuscript. Publishers often concentrate on the job of preparation and typesetting, and think less about the other parts of the process. Librarians worry most about making the literature available to the users. And so it goes. We have analyzed the publishing process and have broken it down into six components. In any publishing activity, all these steps are needed in some form or other, and all must flow easily from one to the other. The six components are:
Manuscripts must contain the logical markup needed to identify all the components, such as title, author, institutional affiliation, abstract, sections, figures, etc. When that is done at the beginning, i.e., by the author, the further processing of the manuscript can be carried out with a fair degree of automation.
When we talk about scholarly publication, some form of review is needed to verify the accuracy, importance and effectiveness of communication. In astronomy, this review almost always results in changes to the manuscript. As we enter the electronic era, there has been considerable discussion about the possibility of having public commentary be posted with the article in lieu of peer review. At the AAS, we have opted not to follow that model, arguing that the people most likely to comment are not necessarily the ones we would choose as editors or reviewers. Under this system, understanding the worth of any given article would require that the reader scan not only every article of possible interest, but also the comments accompanying each article. This makes for more reading at a time when the number of papers keeps increasing.
A consistency of style and clarity of expression is crucial for effective communication. Good copy editing along with consistent and high quality composition and typesetting make an article much easier to understand. These qualities set the better journals apart from other publications. As we enter the electronic era, the AAS has chosen to continue to emphasize quality.
The heart of an electronic publication is the database of electronic manuscripts. A good manuscript database is coded so that different versions of the article can be extracted as needed. While they may contain the same material, the screen version will be different from the hard copy printed locally, which will, in turn, be different from the version which might be put on a CD-ROM. The primary database has to be constructed such that all the different versions can be produced automatically upon demand. But, what is even more important, the database of manuscripts has to be structured so as to be searchable by the reader, and must provide a standard interface for interoperability with other collections of literature and data. The best papers in the world are of no use if they can't be located and accessed when needed.
The production of a major scholarly journal is an enormous job. The Astrophysical Journal alone contains 25,000 pages each year; this translates into over 110 typeset pages every working day. Producing this amount of material requires a well designed system that requires minimal human intervention beyond the copy-editing stage. The production pipeline has to flow smoothly from one stage to the next. The input format has to be designed to facilitate this flow, and the steps at each stage must be well coordinated.
Once the journal is out to the readers, it is tempting to think that the responsibility is over. No so! The scholarly literature defines the corpus of knowledge about any given subject. The publishing job is not complete until survival of the literature is assured. During the last five centuries, libraries have provided the means for storing scholarly literature and making it accessible. In the electronic era, the manuscript files have to be tended both to ensure their survivability into the distant future and to verify that the original versions of the papers are not changed, updated, or tampered with without the reader being made aware of the changes.
In beginning the journey into the electronic future, the AAS has chosen to function in an experimental mode. The world has been dealing with movable type and paper printing for over five centuries. In the rapidly changing era of electronic networking, we believe that only by trying various approaches and seeing how they work will we be able to serve the needs of the users. We expect to modify, enhance or even abandon a given approach depending upon how useful it is to the community in their work. Whenever possible we have taken small steps, embarking on small projects to give us the experience in using the new electronic capabilities.
Another underlying principle has been to recognize the benefit to adopting a broad approach which makes use of a wide variety of expertise. The development of the Electronic ApJ Letters has been accomplished through building teams and coalitions to bring a wide variety of expertise to bear on the problems to be solved. Our effort has been a close collaboration of scientist authors, foresighted publishers, computer networking experts, librarians and the AAS Executive Office staff.
With this background in mind, the AAS developed a plan in 1992 to use the emerging electronic network capabilities to enhance the communication among astronomers (Boyce, Dalterio, & Biemesderfer 1992). With the plan in place, the AAS Publications Board and Council adopted the following working guidelines:
The ``publishing'' enterprises of the Society encompass more than the scientific papers published in the formal literature. As the AAS embarked upon the first experiments with submission and distribution of meeting abstracts, we quickly became aware that the same techniques, operating procedures and concerns apply to all the information and data which one distributes electronically. The AAS now distributes and accepts a wide variety of information of interest to the North American astronomical community. These include:
One immediate lesson we learned was that you can not go partly electronic any more than you can partly jump out of an airplane. What do we mean by this? First, as soon as you offer one set of information electronically, the users will demand that everything you produce be offered in an on-line version.
Second, it is not effective simply to take the paper versions of your material and deliver it electronically. While this is possible, the paper material is often not suited for easy reading on the screen. Nor is the material prepared for paper organized to take advantage of the additional capabilities and hypertext linking which is possible in the electronic environment. An effective electronic presentation has to be designed from the beginning for that purpose. Further, if you plan to collect money, you add another set of complications and accounting procedures which have to incorporated into the system. Going electronic involves everyone in an organization.
The most important lesson we learned is that if the initial preparation of the material is not done with the rest of the process in mind, it can be extremely difficult (and expensive!) to produce an electronic version. Each step of the process has to flow smoothly into the next as automatically as possible. This means that the form and embedded coding in a document must be well defined at document preparation time. Each individual, group or department that handles an electronic document during preparation must understand that they are part of a complete process and that they have to adhere to certain standards if the whole process is to flow smoothly.
While manual intervention is expensive in terms of time and resources, we also discovered that manual intervention is often necessary. Authors make mistakes, or don't follow directions, or want to try something outside the bounds of what our system can accept. Correcting these documents can cost as much in time as processing three or four well-formed manuscripts. Often, it is easier and more cost effective to start from the beginning and retype a document from the paper submission than it is to fix a badly prepared electronic document.
Another lesson that came from our first three years of electronic publishing experience is the need to maintain the on-line database. A document is only useful if it can be found. This means that documents have to be designed to be part of a database which can be searched effectively. Titles, keywords and even abstracts do not provide enough material for intelligent search systems to work effectively. Some documents are relatively static, but others need updating at frequent intervals. Readers know that paper documents cannot be updated frequently. However, an electronic document is expected to have the latest information. Anyone offering material on the WWW must plan for periodic maintenance of their documents. If you are not careful about how much material you offer on the WWW, you may find that the maintenance chores have grown far beyond your initial expectations.
Starting in July 1995, every part of every article in the Letters section of the Astrophysical Journal (EApJL) has been available over the WWW. The development of the presentation and the features incorporated in the electronic version has been accomplished by the AAS electronic publishing team over the course of the last two years. Two full scale demonstrations of sample issues were made at AAS meetings and the feedback from scientists, both at these meetings and over the WWW have been instrumental in the choice of features, presentation schemes and navigational tools. The first production version of the Electronic ApJ Letters has been received favorably by the community, and a new version is due in January 1996.
Aside from the inclusion of all complex math and figures, there are several features which, to the best of our knowledge, first appeared in the AAS journal demonstrations and which set this journal apart from other Internet journals.
The paradigm shift is upon us. The method of communication in astronomy has completely changed in the last five years. Virtually all personal communication is now done electronically. The more formal communication is headed the same way. For instance, all but one or two abstracts for the AAS meetings are being submitted electronically, and appear on the WWW well ahead of the meeting date. Now the journals are beginning to be produced in electronic form.
The future is becoming apparent. We already have observational data from space missions available electronically in the form of spectra and calibrated images. These are the starting points for many research projects. The on-line journal articles are the ending points of our research efforts. But, there is a wealth of intermediate material which will also need to be available on the net: there is the underlying data from which published graphs are made; there are tables of calculations or laboratory measurements, observing lists, catalogs, etc. All are necessary tools, and the users will demand that they be made available electronically. In order to be useful, such collections will have to be available, retrievable and provided in a form which is easily used. We are moving from a situation in which there have been independent collections of data into an era of distributed and interoperable sources of data, of literature, and of intermediate material.
One can no longer simply think about one's own information, but must consider how the data will interact with the other material needed for effective research use. Our vision has to expand beyond our own department, or our own institution.
Material has to be organized so that it can be located when needed. Metadata (descriptions of collections) have to collected, and metadata-bases must interoperate globally. We have to realize that the material we produce and the material needed by astronomers in their research extends beyond the strict boundaries of astronomy. We have to consider how to cross discipline boundaries.
But not all material is being made available electronically at the present time. There is a wealth of material, for instance, most of the ground-based observations, which has to be put on the net if we are truly going to move into the electronic era. While it will be expensive to do so, we believe the community will come to expect to find a much broader range of materials on the net, available for use virtually instantaneously.
The development of the EApJL has been supported in part by The National Science Foundation under grant AST 94-03835. The authors are grateful to Evan Owens of the University of Chicago Press who brought the EapJL to fruition and to Jim Fullton, Archie Warnock and Heather Dalterio for their efforts on this project.
Boyce, P. B., Dalterio, H., & Biemesderfer, C. 1992, The AAS Three-Year Plan for Electronic Publishing, http://www.aas.org/newsletter/nlinsert/boyce1.html