Table 1

Tag density in Connotea, CiteULike and MEDLINE on PubMed citations

System

N sampled

mean

median

min

max

stand. dev.

coefficient of variation


Connotea per tagging

28236

3.02

2

1

115

3.74

1.24


CiteULike per tagging

45525

2.51

2

0

44

2.16

0.86


Connotea aggregate

19118

4.15

3

1

119

5.14

1.24


CiteULike aggregate

19118

5.1

4

0

74

5.29

1.04


MEDLINE

19118

11.58

11

0

42

5.3

0.46


'N sampled' refers to the number of tagged citations considered. For example, the first row shows the statistics for the number of tags associated with distinct posts to the Connotea service. In contrast, the 'Connotea aggregate' row merges all the posts for each citation into one. In the aggregate cases, the numbers of tags reported refer to the number of distinct tags -- repeats are not counted.

Good et al. BMC Bioinformatics 2009 10:313   doi:10.1186/1471-2105-10-313

Open Data