Ten thousand papers a month

Nick-JonesNick Jones is a second-year postgraduate researcher in the School of Mathematics. His research is rooted in physics and the problem of understanding how a large number of interacting particles behave. Particularly interesting is the case where the emerging cooperative behaviour of the whole system is very different to that of the smaller pieces that make it up.

Most of the days that I spend working on my PhD, I am at my desk (which, incidentally, has a pretty great view over Clifton). There, I’m usually confronted by a web browser with so many tabs open that you can’t read the names. Most of them have the same icon though: the ‘smily face and crossbones’ of arXiv.org – a website where many researchers post preprints of their articles, as well as lecture notes and other such useful things. One thing you realise during your studies is just how much information is out there, and what a challenge it is to find the thing you need: at the end of last year arXiv passed one million new papers since its 1991 inception. Initially, the papers were in one sub-field of physics and there would be about one hundred papers added per month. Since then it has grown to cover physics, mathematics and computer science as well as quantitative subfields of other sciences, with almost ten thousand new papers each month! It is a fantastic resource; with many scientists making their work available for everyone to read, including those readers who don’t have the journal subscriptions that I get through the Bristol library. Other fields have similar preprint servers, and not every physics paper appears on arXiv, but some idea of the quantity of new work can be gained from the graph below. newsubsMost of the growth you see is in adoption of arXiv by a subfield, but note that hep-* (the high energy physics category) has had a relatively constant submission rate for the past few years, so the number you see is probably about the size of the active field. This works out at about twenty five papers per day, so a lot to keep track of!

With so much research being done it can be difficult to stay on top the field your project sits in. One way is to meet people at workshops and conferences (and stay in touch – they might write to you with something interesting!), but if you are at the office you will need to keep track of articles on the internet. If the field is moving quickly, you will have to keep an eye on the preprints in case something you need appears and to make sure you don’t end up doing research that would have been new at the start of the year but has now been done somewhere else. Equally, if there is less activity in your field, it’s easy to stop looking for new publications and then you are guaranteed to miss developments. And of course, these are just the new papers, you also need to know what is already out there! Thankfully, the internet makes life easier, but sometimes a bit of luck helps too. In an effort to make sure I didn’t miss anything new I signed up for the ‘scientific recommendation engine’ Sparrho. You put in keywords that you are interested in and then it will check all kinds of websites (including, of course, the arXiv!) for articles. You then tune it by choosing which of these you consider relevant. After putting in three words that I thought most appropriate for my research project and playing around for five minutes it came up with a paper from 2013 that is exactly what I needed and which I could easily have never found! Since then I have been using it to keep tabs on new papers in my field, although I haven’t had another such breakthrough from (recent) history.