Half good and some not good results in a research paper?

Question

I have made some research about a computer science topic. The initial idea seemed good, but when I implement the idea; this works in some cases and in some others it does not bring conclusive results.

How can I write a paper about this subject in particular? I mean, I do not want to hide the bad results. I want to include them, but I feel scared that my paper could not get accepted because of that.

I was thinking also to submit it in a not so prestigious conference, but also I do not see any advantage on doing so.

Answer 1

It's important to remember that a paper is supposed to make some kind of contribution. It's not merely a report on "things you did". So if you have some good results and some inconclusive results, what's the contribution ? As sr3u says, you need to work out what's going on and come up with some answers. It's entirely possible that you fail in this endeavour and end up with NO paper. That's not fun, but it's something that you must accept as a possibility.

Answer 2

It would be good to dig a bit deeper and analyze why the idea works in some cases and not in some others. More than the fact that your idea seems not to work in some cases, a conference PC may be less than impressed by the fact that you have not worked out all its ramifications and clearly identified the structure of the problem for which it is good. In all CS research conferences and journals of some reasonable quality, it is the responsibility of the author(s) to present ideas as well as to work out their consequences in some depth -- it is in general not acceptable to present a bare idea (or an idea not thought through properly) with the expectation that the audience/readers will take it forward on their own.

Presenting more analyses would also make your submission stronger by adding some other theoretical results to the paper. You could also identify future research directions for addressing the cases not covered by your current work, which could add gravitas to your current paper as well as suggest your own future direction.