On the "persistency" of scientific publications: introducing an h-index for journals

05/23/2017 ∙ by Roberto Piazza, et al. ∙ Politecnico di Milano 0

What do we really mean by a "good" scientific journal? Do we care more about the short-time impact of our papers, or about the chance that they will still be read and cited on the long run? Here I show that, by regarding a journal as a "virtual scientist" that can be attributed a time-dependent Hirsch h-index, we can introduce a parameter that, arguably, better captures the "persistency" of a scientific publication. Curiously, however, this parameter seems to depend above all on the "thickness" of a journal.

READ FULL TEXT VIEW PDF
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.

References

  • (1) Here and in the following, by “papers” I mean proper articles, namely, I did not include, for instance editorial material, corrections, or review papers.
  • (2) Note that a few “anomalous” years, in which the total number of citation is anomalously large because of the occurrence of one or two exceptionally cited papers, are not included in the fit. For instance, the huge number of citations for JCP in 1993 is due to a single, fundamental paper in density functional theory, which has already obtained more than 60000 citations, while the value for PLB in 2012 is mostly due to the two papers describing the discovery of the Higgs boson at CERN.
  • Hirsch (2005) J. E. Hirsch, Proc. Natn. Adad. Sci. 102, 16569 (2005).
  • (4) Since we consider only “proper” articles, I did not include in the collection those journal mostly publishing review. I also excluded journals belonging to the Nature group, which show much higher values of the ratio .
  • Piazza (2015) R. Piazza, Europhysics News 46, 19 (2015).
  • (6) More precisely, by examining the individual citation reports of 470 authors of PRL, I found that their -indexes (spanning on the overall the range ) is quite well fit by the relation , where is the total number of paper they have published. This formula allows to predict the -index of a colleague to within 20% when his/her is known.