4

So I want to go about and compare different research groups for their standing, their potential for an exciting and fruitful environment etc. My field is Software Engineering with a heavy tendency towards AI in Reinforcement Learning. I'd prefer to stay in mainland Europe, more specifically the EU until the dust has settled on Brexit.

I googled a few times and found some answers. But as one person correctly noted, that is highly biased. I read "the best" several times and there obviously cannot be 5 number 1. While I don't need "1" everyone is obviously aiming to optimize their future and so am I.

In the past I have been looking at Barcelona, Brussels, Berlin, Zurich, Amsterdam, Vienna, Munich, Karlsruhe, ...

Finally, I decided to go about it a little different and download a few papers to base my decision on the publication / research level of various institutions. So I got myself scrapy and downloaded 2600 papers from ECAI, IJCAI and AAAI. I got them all stemmed and indexed with recoll and thought about doing some kind of structured analysis on them to get an overview of citations, referencing, publication count, names, institutions etc.

Now before I spend a day on this: Has anyone else done this before? Is there some tool that helps doing such analysis already? And is it an appropriate way to get the quantitative side of the research quality of various groups?

  • 1
    There are also some AI research groups in Paris & near Grenoble & Sophia Antipolis (all these in France) – Basile Starynkevitch Jun 13 '18 at 17:25
  • 5
    It's my opinion that your relationship with your supervisor and colleagues will have a bigger effect on the success of your PhD than the output of the group you join. Find someone to work for that you get on with and the rest will follow. – astronat Jun 13 '18 at 17:35
  • @astronat I agree that the people matter greatly. But obviously, if you cannot meet and visit all of them, prioritizing the list based on information you can accumulate now and then pick the group that also clicks from a social level is a good approach imho. – pascalwhoop Jun 13 '18 at 19:19
  • 1
    @Hello THAT is exactly what I am trying to avoid. Some random people giving me their point of view. Unless I get a decent sample of some 1000+ opinions, it's inherently biased. There are equal initiatives for Cologne, Berlin, Paris, any major city or region really. The question is how to create a sorted priority list from these based on some metric that is measurable and fairly representative. – pascalwhoop Jun 14 '18 at 12:24
  • 3
    It is a mistaken belief that statistics solves such a individualistic problem. Statistics are good if you work with unnamed numbers of instances and all you are interested in are averages and variances thereof. But if you are going to have a one-shot (or, if you are unlucky, two-shot) attempt at a PhD, then you need to collect and evaluate qualitative rather than quantitative data. Statistics will only tell you so much. – Captain Emacs Apr 7 at 19:09
1

This answer is coming very late, but just in case someone finds it useful:

http://csrankings.org/

This site ranks universities based on their research output in select top conferences in various fields, you could use this to narrow down your list.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.