Transcript Slide 1
The Publication of Research Information: Does Astronomy’s Present Portend a More General Future? Michael J. Kurtz Harvard-Smithsonian Center for Astrophysics The Current State • • • • • • • 10,000 professional astronomers $6,000,000,000 yearly budget ($600,000 per astronomer-year) 3½ Large Main Journals + ~5 Smaller High Quality ~8000 High Quality Papers/year ($750,000 per paper) The Current State – Also • High Quality Articles from Physics and Planetary Science • Several Dozen 2nd tier journals • Many Conference Proceedings • Many Abstract Books • 49,000 total papers/year • 17,000 refereed in some manner Besides the Journals • • • • ADS CDS Data Archives (Virtual Observatory) arXiv Astronomer CDS ADS Data Centers Journals Data Links Data Links arXiv A Quick Tour of ADS • 1st Find all papers in refereed journals by me which have the word “halo” in the title • 2nd Get the citing papers for one of these papers • 3rd Notice the equivalence of arXiv eprints and journal articles in the ADS system ADS Tour - Part Two • 1st Use combined CDS-ADS search to get papers concerning the x-ray properties of the Cluster of Galaxies Abell 754 which have Data Links, sort by citations • 2nd Notice that all the most cited papers concern older experiments ADSTour – Part Two Continued • 3rd Use data mining techniques to discover what papers are currently being read by people who read the 79 papers on Abell 754 x-ray properties with Data Links • 4th Notice that these most popular papers have Data Links to the most current data sets on the x-ray properties of clusters of galaxies The Publishers Nightmare? • myADS-arXiv is a fully customized to each user, open access virtual journal which includes nearly every important paper in Physics and Astronomy about four months before it is available from the journal. The Effect of the arXiv • What is the readership pattern for articles in the arXiv, how does that differ from the journal version of the same article? The Effect of the arXiv • What is the citation pattern for articles in the arXiv, how does that differ from the journal version of the same article? The Effect of the arXiv • Currently 75% of Astrophysical Journal articles are preprinted in the arXiv. What is the difference in citation histories for those which are preprinted compared with those which are not? • What is the readership history? The Effect of the arXiv • How has the arXiv changed the citation behavior of astronomers? The Effect of the arXiv • The citation advantage of being in the arXiv is also present for Physics • The time history of citation impact factor for Physical Review D (nearly 100% in arXiv) shows this The Effect of the arXiv • The difference in change of citation impact with time for Physical Review D (~100% arXiv) with Physical Review B (~40% arXiv) also shows this • As does the difference in citation impact with time for Nuclear Physics B (~100% arXiv) with Nuclear Physics A (~35% arXiv). Conclusions - Comments • With a total cost of <1% of the cost of personnel the journals are cheap. Far more important than cost for astronomers is utility. • It is not possible to be in a position to write an article for the Astrophysical Journal without also being able to read the Astrophysical Journal. • Physical Review also. Conclusions - Comments • Astronomers want to read the articles ASAP, they are quite willing to short circuit the refereeing process to do so. • Astronomy’s preprint culture has led to a situation where an article must be put up on astro_ph in order to be noticed, just being published in the journal will not do. • Also true for Physics Conclusions - Comments • The danger in the current situation is to the brand. When all the important articles are in astro_ph, and people say they saw the article in astro_ph the value of a brand, such as The Astrophysical Journal is diminished. • This may well kill some of the weaker brands very quickly. Conclusions - Comments • The solution is for journals to put up there own “preprint” versions, either after submission or acceptance, using a name (doi?) which always provides a permanent link to the current version.