The problem has been asked before, but under a slightly different conditions.
There are acronyms, that have multiple meanings based on the context, like for example in this question wai which can mean w3c-wai or haskell-wai. In such cases, the solution is quite simple - we can retag questionsso they are more specific and not use wai at all (mark it as "deprecated"). But in some cases, adding any additional information to the tag discards the actual justification of acronym usage in the first place.
machine learning, it's common acronym is
ml (or even more common
ML). In the same time, on stack overflow we have ml which points to the
ML is a family of functional programming languages, created by the Turing awarded computer scientist Robin Milner. It was initially created as the metalanguage for a theorem prover (hence the name), but quickly became used as a general-purpose programming language. One of ML most famous characteristics is type inference supporting parametric polymorphism.
Of course we could create artificial tags like ai-ml and language-ml, but using acronyms is justified by their simplicity, and I don't think that actuall anyone asking about machine learning would come up with tag ai-ml (or ml-ml ;)). I currently simply retag questions with ml in cotext of machine learning to machine-learning, but I don't think it is isolated issue.
Is there an already developed way of dealing with such phenomen? Or is it too rare to be considered?