Researchers in our area are confronted with two goals that easily conflict with each other:
The best way to deal with this situation is not to abstain from putting articles on line, but to be informed about the rules that apply and to use them in the best possible way. Each author has to look out for his or her own interests.
As a help to researchers, we have put together a summary of the explicitly specified rules of three professional organizations, namely IEEE, ACM, and AAAI. The summary of their rules is found in a special overview page. We will extend that overview of current rules as soon as we get access to rules by other similar organizations, or when we obtain clear policy statements from commercial publishers, including of course organizations in as many other disciplines as possible.
As a further help to researchers, the present page and its sub-pages describe a few alternative, reasonable procedures for how to deal with a research article before, during, and after its reviewing and publication by a journal or conference. The primary consideration in these procedures was that they should comply with the policies of the three organizations mentioned above, as well as with what we know (at least implicitly) about the policies of other publishers. Furthermore, of course, the procedures must be reasonable in their own right, from the point of view of both authors and readers.
For the Latex style that is used by the E-Press series in Computer and Information Science, etendu, we have also developed special Latex commands which will format the first two pages of the article in conformance with these respective procedures (only partly ready). Similar support for use in other text preparation systems (Framemaker, Word) would be most welcome.
The advantage with article locator pages is that they allow one to write citations using a fixed "address" through which the article can always be found, but at the same time it is possible to re-link to new versions of the article and to new places where it may be stored.
The Linköping E-Press uses a variant of ALP:s for every work that it publishes, but it is proposed within IDA to also set up an ALP register for Computer and Information Science, and to use it more freely than in the E-Press scheme. The ACM Interim Copyright Policy seems to assume the use of article locator pages.
Standard publishing procedure 1 is the cautious one with respect to using the Linköping University Electronic Press. Its basic idea is that the author puts up successive versions of his or her article in his/her own WWW page structure, and uses an Article Locator Page (e.g. as provided by IDA in the proposed new IDA publication system) as a well defined way of obtaining a URL for referring to those article versions. In addition, E-Press publication may be used for specific purposes when clearly allowed by the external publisher of a journal or conference proceedings.
Standard publishing procedure 2 is the more aggressive one with respect to using the Linköping University Electronic Press. Its basic idea is that the author publishes the first version of his/her article in the E-Press as soon as it is ready, and then submits it to journal or conference reviewing afterwards. This is clearly acceptable for some journals/conferences; it is likely to work fine for workshops in most cases; it is the only mode of operation for the new European AI publishing system (Electronic Transactions on Artificial Intelligence); it is almost certainly OK with IEEE and has a good chance of being OK with ACM, and it is almost sure to preclude external publication in some journals or conferences.. The user of this procedure must therefore be careful to know what he/she is doing.
Each of these alternatives has its pros and cons. With the first procedure, you avoid getting into trouble with the external publisher, but you expose yourself to the risk that's involved in putting your material on-line on the Internet without proper protection: you are not able to cite this work as your own during the period that it is posted but unpublished.
For the second procedure, the risks are reversed: you have at least a nominal protection of your results since it is definitive that the work has been published, but in adverse cases you may lose a chance of getting the work republished in a refereed periodical. Notice that in this case it is recommended to "promote" your article and make people aware of its existence, for example using mailgroups or newsletters: only publishing it in the E-Press does not bring it to the world.