Reference requests and the like

Question

Someone recently asked a question about references for work in pattern recognition with fuzzy logic. Leaving aside the issue of topical scope, here's the (harsh) comment I wrote:

I apologize if this sounds rude, but it's not really our job to help you do a literature search, especially since you haven't provided any evidence of prior efforts in this regard. What you're asking us to help with is the kind of thing you should learn to be doing: navigating an area, finding out what's important (not what's hot!), and so on. If, after doing so, you have a specific question, then that might be more appropriate.

This might be my own impatience, but I'm beginning to see more questions of the form of "tell me what what's known about area X, or what's the latest in area Y, and so on", without any evidence that the OP has done some research, and has genuinely hit a brick wall of the form "This specific question appears to be very natural, but I can't find anything on it".

I want to encourage people to ask SPECIFIC questions and not go on fishing trips and expect the community to help out - I'm not a Google replacement.

Am I being too harsh ?

Answer 1

First of all, I believe your comment is not harsh, yet it can be improved to become a friendly advice. In addition, I believe you must be a moral person to ask this question.

Unfortunately, it is very hard to know what to do when things happen, since we don't have a precise rule that distinguishes "right" and "wrong" questions (and, IMO, such rule is not conceivable, simply because we do not live in a 0-1 world).

This community is like a newborn infant. It takes time to develop, and till then, we must be patient. We must treat newcomers with patience, and teach them how to ask a good question. Harsh behavior results in a bad first impression, and they may go and never come back.

I suggest tolerating wrong questions at least while we are in the beta phase (and maybe for several months after). IMO, the following scheme works best:

1. Be suggestive. Point out where the problem lies, and how to restate the question to make it worth answering.
2. Advise the user to read part of the FAQ, or refer him/her to a link in meta. Be as friendly as possible, since newcomers aren't familiar with the rules.
3. Wait for a while (say, several hours or even a day) and see if he/she corrects the problem.
4. Down-vote or vote to close.

Answer 2

My personal feeling is that if you are asking for directions to the literature or for a specific reference, you owe it to the community to give something back by explaining the problem and maybe teaching a little in the process. That's what I tried to do in my two questions so far. If something like this isn't done, the site risks becoming less of a community and more of a mechanism for one-to-one exchanges of information.

In the case of an undergraduate seeking guidance, this is maybe asking too much. But at least he could try to explain something about his background and his interests in the area, so posters don't waste their time writing up irrelevant answers.

Answer 3

I think that this is an important discussion to decide in which direction our website should go, so thank you for raising this issue. I do not have an answer, but here is what I thought about this specific question.

On one hand, I agree with you that the asker was just asking for a broad literature search to get accustomed to a field, and the asker has not shown the evidence that he/she has tried.

On the other hand, I see some inconsistency when I compare this question to the question asking for ideas of graduation projects about quantum computing, which has been perceived positively. One is about the literature search and the other is about project proposals, but this difference does not seem important (if anything, asking for the literature is more specific). Note that the question about quantum computing projects does not show any evidence of prior efforts, either. Moreover, I feel that the topic “applications of the fuzzy logic to pattern recognition” is more specific than “quantum computing” (I may be biased here; quantum computing is my field of research, and I am not familiar with either the fuzzy logic or pattern recognition). I have a difficulty justifying the difference of people’s responses.

Added: Just to make sure, the question on quantum computing projects I am talking about is not the same as a more recent question about how to write a research proposal on applications of evolutionary algorithms to quantum computing. The latter seems to be of a different nature (therefore I will not talk here about whether it is appropriate or not).

Answer 4

My feeling is that it is perhaps a little harsh. My own questions have tended to be quite short and simply ask for the current state of knowledge on some topic, and each has been significantly up-voted. Perhaps I should be asking different questions or going into more detail, but my feeling is that simple questions are easiest to answer.

I know that there are a fair number of junk questions, but I think it is important at this early stage to keep the site as friendly as possible.