This week I dinged a thousand bookmarks on del.icio.us and thought it might be interesting to examine how I’ve used the service over the last couple of years. Firstly, some raw data:

  • I’ve posted 1000 bookmarks since 25th March 2004. That’s 1480 days ago, so I posted an average of 4.73 bookmarks a week.

  • Amazingly, I used 999 distinct tags, so obviously I’m introducing an average of one new tag for every bookmark. 568 tags are used only once. The most-used tag is design, appearing on 104 bookmarks.

  • The number of tags on individual bookmarks ranges from 1 (on 11 bookmarks) to 15 (on 1). There are 4702 tags in all (including duplication across different bookmarks), and so each bookmark has an average of 4.7 tags.

I guess the most noteworthy thing out of all of that is the coincidental near-exact matching of the number of bookmarks with distinct tags (1000/999), and the average bookmarks per week with average tags per bookmark (4.73/4.70). I have no idea if this really is just a coincidence or an effect of some unconscious underlying usage pattern.

Here’s the frequency of my posting those thousand links:

![Graph of number of bookmarks over time](http://chart.apis.google.com/chart?cht=lc&chs=635x300&chds=0,80&chd=t:3,0,0,0,0,0,0,4,1,0,1,0,62,59,27,25,54,56,67,52,41,21,30,37,47,27,24,19,14,30,17,6,24,17,19,34,8,24,22,15,17,8,21,12,11,7,15,12,10,4&chxt=x,y&chxl=0: Mar+2004                 Jan+2005                       Jan+2006                       Jan+2007                       Jan+2008     Apr 1:   40+posts 80+posts)

Here are my top 10 tags of all time:

![Bar chart of top 10 tags](http://chart.apis.google.com/chart?cht=bhs&chds=0,110&chs=635x240&chbh=16,5&chco=4d89f9&chd=t:62,67,69,73,74,80,81,81,87,104&chxt=x,y&chxl=%200: 0 10 20 30 40 50 60 70 80 90 100 110 1: design programming urban art mapping webdev photography linklog opensource psychogeography)

But really, the commonly used tags are only a small part of the data compared to the long tail of little-used tags. The same top ten tags are to the left of the vertical blue line on this chart showing the number of times each tag was used:

![Distribution of tags](http://chart.apis.google.com/chart?cht=lc&chs=635x200&chds=0,110&chco=2CBD33&chd=t:104,81,81,73,67,58,54,51,50,49,47,40,35,33,31,28,26,26,25,24,24,23,23,22,21,21,20,20,19,17,17,17,17,16,16,15,15,14,14,13,13,13,12,12,12,12,11,11,11,11,10,10,10,9,9,9,9,8,8,8,8,8,8,8,7,7,7,7,7,7,7,7,7,7,6,6,6,6,6,6,6,6,6,6,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1&chxt=x,y&chxl=0:   1:           50+posts           110+posts&chm=v,4d89f9,0,5.0,1.0)

There’s lots more that could be read from a dataset this size; it would be nice if del.icio.us actually did some of these calculations for you. I made these charts with the del.icio.us API, Google Charts API, and the handy Python Google Chart wrapper. Congratulations to Mr. Chuck Klosterman of Esquire Magazine (whose book I’m also reading right now) on being my 1000th customer.

— 14 Apr 2008