IJCAI Policy on Multiple Publication of Papers

In resolving this potential conflict, the notion of an archival publication seems central. It seems generally accepted in scientific circles that a particular scientific achievement should be described in a single definitive paper and that this paper should be published in only one archival publication.2 Presumably, this restriction prevents researchers from pretending to a greater scientific output than they have achieved and saves their colleagues from wasting time reading about the same achievement more than once. After some discussion, the IJCAI trustees decided that the IJCAI proceedings is an archival publication and that it should, therefore, only accept papers which had not already appeared in an archival publication. We also implicitly assumed that all AI journals and the proceedings of AAAI conferences and similar national and international AI conferences were also archival publications. Thus, we would reject from an IJCAI conference, any paper that had already appeared, or was about to appear, in another archival publication.