21
$\begingroup$

As a Theory B person, I feel the site really lacks Theory B questions. To make matters worse is he previously active Theory B people are now much less active. More generally, the questions here seem to be mostly in fairly small areas and not sufficiently covering TCS as we all see it.

What can we do to expand the scope of cstheory and attract theoretical computer scientists whose focus goes beyond the questions that we see these days? More generally, how can we broaden the range of questions that appear on cstheory.stackexhange.com?

(Forked of from this more general question.)

$\endgroup$
  • 2
    $\begingroup$ It's interesting to note that most of the high rep users are computational geometers or complexity theorists. Even within theory A I don't think there's a uniform distribution of people. $\endgroup$ – Robin Kothari Oct 6 '10 at 12:22
  • 2
    $\begingroup$ And let's not forget that there is a lot of TCS outside theory A and B... $\endgroup$ – Jukka Suomela Oct 6 '10 at 14:09
  • 1
    $\begingroup$ Theory C and D, you mean? $\endgroup$ – Dave Clarke Oct 6 '10 at 14:11
  • 2
    $\begingroup$ What is this random assignment of letters all about, anyway? It is related to computer science? It has been expressed formally and proven? Great, call it TCS! $\endgroup$ – Raphael Oct 6 '10 at 22:10
  • 1
    $\begingroup$ @Raphael: See this question about the letters cstheory.stackexchange.com/questions/1521/…. These days ICAPL has a Theory C track (Security and Cryptography Foundations) and Theory D is something I introduced to capture (Misc). $\endgroup$ – Dave Clarke Oct 7 '10 at 6:43
  • $\begingroup$ Actually nowadays ICALP C is "Foundations of Networked Computation". Anyway, it varies; we can safely use C as catch-all, no need for D. :) $\endgroup$ – Jukka Suomela Oct 7 '10 at 7:29
  • $\begingroup$ @Robin: questions and answers in computational geometry and complexity have been the ones garnering lots of votes. That doesn't necessarily mean the people who made those posts identify with those labels. For instance, Noam Nisan is often seen as an AGT person, yet there hasn't been that much AGT here yet, and some of the other high-reputation people seem to be in algorithmic graph theory, programming languages, and security. What we perhaps need is more people outside CG and complexity to vote, so that other areas don't get neglected. $\endgroup$ – András Salamon Oct 8 '10 at 0:15
  • $\begingroup$ I'm curious if there are many people with research interests both in Theory A and Theory B. If there are such people, they might join b/c of Theory A but then start asking Theory B questions. $\endgroup$ – Lev Reyzin Mar 8 '11 at 16:42
  • $\begingroup$ Wasn't there already a discussion about this started by Joel Spolsky? The big kahunas laid down the law against this kind of expansion. Here's the link (meta.cstheory.stackexchange.com/questions/947/…) $\endgroup$ – davidk01 Mar 18 '11 at 18:03
  • $\begingroup$ @davidk01: The question you commented on predates Joel Spolsky's question. The current question talks about a strategy for bringing more CS Theory people on board. $\endgroup$ – Dave Clarke Mar 18 '11 at 19:16
  • $\begingroup$ @Dave Clarke: Yup, I got my times mixed up. $\endgroup$ – davidk01 Mar 18 '11 at 19:49
17
$\begingroup$

This is a list of questions at CSTheory focused on Theory B (Logic, Semantics, Automata
 and Theory of Programming).

This list is posted on the principle that one way to attract people is to highlight questions and answers related to their interests. This is a community wiki; please edit the list if notable new questions are posted.

$\endgroup$
  • $\begingroup$ This is certainly a valid way of attracting people. But how do we get them to visit and look at this list, short of spamming people? $\endgroup$ – Dave Clarke Feb 16 '11 at 15:16
  • $\begingroup$ I ask nicely at lunch before sending the email. That may still qualify as spamming, though. $\endgroup$ – Radu GRIGore Feb 16 '11 at 15:19
  • $\begingroup$ I keep bugging my friends until they show up. That certainly is spam. $\endgroup$ – Raphael Feb 16 '11 at 15:47
  • $\begingroup$ I was hoping to see a list like this! I need to email Andy Gill a blurb to go in SIGPLAN Notices, and needed a good list of questions. $\endgroup$ – Neel Krishnaswami Feb 16 '11 at 16:28
  • $\begingroup$ I think Kaveh's idea below about retagging is also a good one. $\endgroup$ – Suresh Venkat Feb 16 '11 at 19:57
  • 3
    $\begingroup$ Maybe we could get Phil Wadler to include a link in his blog. Or perhaps a mail to the types mailing list. $\endgroup$ – Dave Clarke Feb 17 '11 at 7:47
  • $\begingroup$ Phil Wadler has a blog ? $\endgroup$ – Suresh Venkat Feb 17 '11 at 22:55
  • 1
    $\begingroup$ Yes. And sometimes it has nice crocodiles. wadler.blogspot.com $\endgroup$ – Radu GRIGore Feb 17 '11 at 23:00
  • 2
    $\begingroup$ @Dave: Also, spamming seems to work best if you send a focused email: pick one question and send the email to someone you know will be interested. ;) $\endgroup$ – Radu GRIGore Mar 3 '11 at 19:32
  • 3
    $\begingroup$ See wadler.blogspot.com/2011/03/theory-b-questions-at-cstheory.html for a (crocodile-free) pointer to here. $\endgroup$ – András Salamon Mar 6 '11 at 13:10
6
$\begingroup$

Part of the problem can be tagging: Theory B questions are not well tagged. It would be nice if some expert in Theory B goes over the questions and adds more tags to those questions (at least one top level tag + as many as possible relevant tags).

(This may annoy some users since it will bump up old question to front page, I guess a less aggressive retagging (i.e. retag a few questions each day such that they don't fill the front page) would be more suitable.)

$\endgroup$
5
$\begingroup$

This is not an answer but an obvious observation, but I post it anyway in the hope that stating the obvious may shed some light on the situation.

Honestly, I think that the situation is pretty much chicken or the egg. To begin with, people (not necessarily in Theory B) visit the website because people are already here. If the pioneering people had failed, we could have ended up with an empty restaurant — no one is there, therefore nothing interesting is there, therefore no one wants to visit there.

Thanks to the first participants (especially those who participated while the site was in private beta), the website as a whole seems to have survived the risk of the empty restaurant syndrome. However, that does not mean that the risk has gone in each subfield of the theoretical computer science. I see many strong questions in complexity theory and algorithms, so I guess that the website has survived the risk of the empty restaurant syndrome in these areas. However, in other areas, I think that the website is still in the middle of the risk.

$\endgroup$
  • 2
    $\begingroup$ I agree. I am involved in the model checking community (less than a few years back, but still) and my opinion is that (1) there is not enough activity here concerning this field to attract big names and (2) this community is half TCS and half Engineering. So, to attract them, question must be more motivated (that's my 2 cents, no more). $\endgroup$ – Sylvain Peyronnet Oct 7 '10 at 8:22
5
$\begingroup$

Regarding theory B, the problem is one of promotion. I don't know too many people in that realm, and so I can't even go around emailing people asking them to participate. I pinged Luca Aceto (because I know of him from his blog) asking if he'd mention the site, and it may have slipped his mind. Of course even if he did mention it, there is the empty restaurant problem that Tsuyoshi describes. I always make a point of upvoting questions and answers in that realm if they make any kind of sense, but that's not enough.

For theory B to increase in participation, we need help from theory B folks already here to make a concerted push and get other folks to come and ask/answer questions. It'll take some time, but they should be able to acquire rep quickly. For example, one question on functional programming made it to Reddit and skyrocketed up the charts.

The same is true for other areas of theory A. For example, approximations, data structures, randomization and probability are areas that are heavily represented in the community but not here.

The best bet really is relentless promotion in forums where the people you're looking for show up. For example, target social events or other kinds of activities at conferences, or even request 5 minutes at a bisiness meeting to promote the site. In fact I was planning to post another question soliciting ideas for what to do at FOCS (I'll be going this year). Advertise nice questions, encourage people to post, upvote them heavily when they do come, and repeat.

Or come up with a splashy question that draws some attention. I don't know what that is for theory B, but I do think that the big-list/CW questions that we love to hate on meta actually might have one benefit in that they allow even marginal participants to contribute and get some credit (via badges if not via points)

$\endgroup$
  • 1
    $\begingroup$ Perhaps I'll think up a good Theory B CW post to attract some interest from Google and then the world. $\endgroup$ – Dave Clarke Oct 6 '10 at 14:45
  • 3
    $\begingroup$ Another thought. Moshe Vardi is editor in chief of CACM. Maybe a promo piece there ? And if that's too broad, maybe a collective joint article in SIGACT and EATCS ? $\endgroup$ – Suresh Venkat Oct 6 '10 at 16:52
  • $\begingroup$ They are certainly good ideas. The article could even be written on the meta site as a community wiki post. $\endgroup$ – Dave Clarke Oct 7 '10 at 6:58
  • $\begingroup$ yes, that's what I was thinking. $\endgroup$ – Suresh Venkat Oct 7 '10 at 8:19
  • 3
    $\begingroup$ @Suresh You got a point: we need Moshe Vardi, he can ask thousands of interesting question a day, and probably answers as well ;) (BTW the first talk I gave in my career, years ago was at LICS, and Moshe was sitting in the first row, it was excruciating for me ;)). $\endgroup$ – Sylvain Peyronnet Oct 7 '10 at 8:19
  • 2
    $\begingroup$ @Suresh: Maybe we should have something like the private beta (only not private) specifically for new areas we want to see more of on the site. To use Data Structures as an example: maybe we could have a "data structures week" where we really try to get lots of data structures people, questions, and answers. Could there be a badge for participating in data structures week (with some extra points associated with it)? A concerted effort to get the word out on the data structures blogosphere and to our data structures colleagues? etc. $\endgroup$ – Joshua Grochow Oct 11 '10 at 15:17
  • $\begingroup$ The only existing badge is this: if the questions associated with a specific tag that you have answered get 100 votes (your answers that is), you get a badge for that tag. $\endgroup$ – Suresh Venkat Oct 11 '10 at 15:34
  • $\begingroup$ This is a good idea in general though. Needs some coordination. I can imagine approximations, data structures, geometry and ml-theory being good topics $\endgroup$ – Suresh Venkat Oct 11 '10 at 15:36
  • 1
    $\begingroup$ When I asked if there could be a badge, I was wondering if there's some way we (or perhaps more likely you, the moderators) could coordinate with the folks at StackExchange to create a badge for such an event. If it's not too difficult I would think they'd like it, since it's trying to expand the community. I agree it needs coordination in general. $\endgroup$ – Joshua Grochow Oct 11 '10 at 16:10
  • $\begingroup$ it's doable. If you could sketch out an outline of how this would work, we could relay it to the SE admins. $\endgroup$ – Suresh Venkat Oct 11 '10 at 16:19

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .