Well I realized this when I was submitting our short paper for CIKM this year. Apparently, it got in. They dint find it as shitty as I did!
Nevermind, So there is a certain use of language in research, its a language that hides well the reality of our results, experiments and our understanding of the problem and the proposed solution. Its a clever veil to keep others at a distance, so that we do not face the criticism that we know we should!
When authors write:
[Abstract] Our method performs exceptionally, fairly or significantly better ==> Well we all know these adjectives have degrees. Exceptionally means, well we nailed it, we took either such simple baselines or such poor methods that any thought into the approach had to beat the previous numbers. Fairly means, we tried and tried, but our approach works only either for some datasets or by tuning parameters you don’t want to waste time tuning. Significant isn’t an English usage, is more statistical here. Statistical significance is only mentioned because we couldn’t use grand words for our results. We say statistically significant because it was time we submitted this work and it isn’t going anywhere. Plus we are making effort to conduct such tests. You cant reject the paper on ‘No statistical significance checks!’.
[Conclusion] We leave X,Y and Z for future work ==> This paper on its own was a lot to take. We leave the onus on other over enthusiastic students to not only verify our findings but to see the faults we see so clearly that they are willing to write another paper on this subject. The other (and often) explanation == We oversimplified our problem to such a level its not realistic anymore, making it realistic, testing it on real data or with proper model is definitely this guys PhD!
[Experiments] We tuned X, Y Z methods and report results of A, B, C ==> This is actually a warning. We tried bits and pieces, it dint work. And we are already struggling with grammar checks on this paper, we dint have space or time to mention more failures. And we are only mentioning it here so that you know that we know such things can be done (well they dint work!) and we are letting you know with as little words as possible.
[Related Work] Previous work A, B, C is very close to ours in D, E, F ==> This is sweeping things under the rug. And if you happen to review a paper with alot of such statements please do check the D, E and F. They would be probably different just by an inch. This is also an admission of guilt of some sorts. Its like these people built this first, but I find this idea so fascinating that i still work on this, and im just modifying (or borrowing) their idea to do something more insignificant.
[Approach] We assume parameters A, B and C for simplification ==> This isn’t a simplification. Its a complication you have no idea about. Read on and our experiments will tell you in detail about how much pain have these parameters caused. So much for the enthusiasm to join grad school to do research on cool stuff!
I will update these as and when I encounter them or use them in my papers!
What?A Phd! 