What does "tagging" mean in experimental high energy physics? Could someone explain in details the meaning of the terminology "tagging" in experimental high energy physics and how is it used in the analysis?
 A: As @dmckee points out this term is used in multiple contexts, but "b-tagging" is probably the most common usage. As in "the selection requires 2 tagged jets" or something of this sort.
b-tagging is the identification of jets (the spray of color neutral particles originating from a colored quark or gluon) that are likely to originate from a $b$-quark. This can be useful for analyses with a lot of jets, of which some are $b$-jets (it will reduce combinatorial background).
Examples include top-pair production (each top almost exclusively decays into a W boson and a b quark), Higgs (couples strongly to b's), of flavour physics (e.g. B-Physics) A disadvantage is that b-tagging algorithms are not very efficient (common workin points include 50%-70% so out of all true b-jets you will only catch a little more than half)
B-tagging works because the B-mesons, that come early in the hadronization process have a significantly longer lifetime than other mesons. This is due to the $V_{xb}$ matrix elements in the CKM matrix (x being one of u,c,t). A drop down into lower generations via weak decay has a low probability, therfore the lifetime is long.
To find b-jets one possibility is to look at jets coming from secondary secondary vertices macroscopically apart from the primary interactions. Another way is to look at the impact parameters (the distance to the primary vertex) of the tracks of the individual charged particles that constitute the jet.
Aside from b-tagging, there is also c-tagging, since charmed mesons have a significantly larger lifetime as well (compared to so-called "light-flavors" u,d,s), but it doesn't work as well. Collectively the term is "flavor-tagging"
Another usage is when people talk about "tag and probe". This is a method to calibrate and measure the performance of some selection. For example it is crucial to know how well muons are reconstructed. Therefore one will select events from a region in phase space that corresponds to $Z$ boson decays to two muons. One will apply strict requirements for one muon but only loose one to the other. The strictly selected one is the "tagged" muon and the looser one is the muon on which one wants to "probe" the reconstruction efficiency (or something else)
