The number of distinct tags assigned per PubMed citation by the aggregate of CiteULike users. The figure provides a probability density histogram of the number of distinct tags per PubMed citation within CiteULike. For each citation, the non-redundant set of tags assigned to that citation by all of the users to post it to their collections is counted. The peak of the distribution (just under a density of 0.15) is at two tags per citation with nearly equivalent densities at one and three tags per citation. Beyond three tags, the distribution drops off smoothly with densities becoming negligible beyond about 20 tags per document.
Good et al. BMC Bioinformatics 2009 10:313 doi:10.1186/1471-2105-10-313