Friday, December 28, 2007

Trend of ArXiv Submissions

Click to Enlarge

The plot above shows the average monthly submission rates on the arXiv from '91 until December last year. The colors indicate the different areas blue: hep, green- cond-mat, red: astro-ph, purple: math(+pink: math-ph). The clear extensions of the bars indicate cross-links. Note how hep stagnates and drops, while math is taking the lead. Below is the statistics for hep only

click to enlarge

I'd be interested to see the statistics from 2007, do you think this trend continued?

You find more of the arxiv statistics here.

This post is a latecomer to our A Plottl A Day series.


CapitalistImperialistPig said...

Holy shit! Is it really fifteen years since I started reading Arxiv? Ouch! I really am old.

Anonymous said...

This is very easily explained: virtually all papers in high energy physics appear on the arXiv nowadays; whereas there are still many subfields of math where substantial numbers of papers are not posted to the arXiv. As these subfields saturate, eventually mathematics will level off as well.

Since the aeXiv classifies subfields of math, it might be interesting to see the statistics broken down by subfield and see whether the different subfields are saturating at different rates.

Peter Shor

Bee said...

Hi Peter:

What is 'easily explained'? I was not wondering about the continuing rise in math and cond math, and a leveling off in hep, but about the drop in hep, and the fact that there are in total more submissions in math-ph. Also, one has to further take into account that there is a general increase in the community (as well as in the world population overall), which should underlie the trend. Plus it might be that in North America/EU most of hep posts on the arXiv, but the rest of the world is still catching up. I.e. you might have noticed in the last years the increasing amount of submissions from the middle east. Taking all this together, how is the hep drop 'easily explained'? (It isn't much of a drop though, so maybe just a fluctuation). Best,


Bee said...

I mean cond-mat. Also, I should have added, I actually think people publish too much, so I find this trend a good development. Best,


Dylan Thurston said...

Perhaps one explanation for the drop in hep-th is that some papers that were posted there were somewhat misdirected; for instance, I know some mathematicians who posted there before switching to the math archives. On the other hand, the timing is wrong for that; I would expect that shift to have happened around '96 or so. But maybe there were other shifts from hep-th to other archives.

Dr Who said...

There seems to be something wrong with either the plots or my interpretation of them: on most days, astroph papers outnumber hep-th by a factor of at least 2, right? It's not unusual for astro-ph to have 50 papers on one day.

Anyway, I am sure that the number of hep-th papers has gone up in 2007. You used to see 12-15 papers each day; nowadays 17-20 is common. No comment as to what has happened to quality.......

Bee said...

Hi Dr. Who,

This was also my thought, but I think the reason why astro-ph is so crowded is that it contains all shades of experimental papers as well, which hep-th doesn't. The blue 'hep' bar above is hep-th/hep-ph/hep-lat and hep-ex. Which taken together might very well be about the same amount as astro-ph (can't tell, I hardly ever read hep-lat and hep-ex). It's definitely overdue to split the astro-ph arXiv. I think I read somewhere that it's actually planned, but not sure what the status is of that.



Bee said...

Hi Dylan:

That's a good point actually. It seems to me in fact that many papers that earlier ended up on hep-th or -ph are now better classified as general physics or history of physics etc. Also, it might be part of the submissions went into quant-ph and gr-qc, which in '06 seem to have constituted about 9.5% (see website I linked to above), but don't know what the trend over the last year was. It could in fact be that esp the quant-ph community has recieved an increasing amount of attention. Best,


Alejandro Rivero said...

Look the current picture (2011)... Really math has taken the lead!