Order and consistency in tagging

Question

This post is primarily to get a consensus on appropriate tag name conventions. Signal processing is a very diverse field and not everyone might be familiar with all acronyms (other than basic ones). I see an increasing trend of acronyms for all tags and it could get confusing eventually. Personally, I think descriptive tags are better than acronyms (which should be synonyms). However, we'd also run into tag length issues (e.g., discrete-time-fourier-transform is 26 chars and only 25 are allowed).

So my suggestion is to use a hybrid of partial acronyms and partial descriptive names as in kernel-pca, discrete-time-ft, discrete-ft, continuous-time-ft, etc. when there is one among the group that will overflow and fully descriptive names in other cases as in hidden-markov-model, fourier-transform, cosine-transform

Answer 1

It all comes down to what people will be expecting these tags to be. For example, when I ask a question related to the discrete Fourier transform, I will start typing dft and expect it to be there. I think that if we go as far as creating full descriptive tags, we have to make sure they all have synonyms among the acronym style tags. With hybrid acronyms, it would sometimes be especially difficult to predict how the hybrid acronym was constructed to start with. In any case, how would you create an hybrid acronym for something like multiple-frequency-shift-keying? I think synonyms is the answer, and several people should really be monitoring how well this structure is formed at the initial stages to avoid confusion.