POS tagging is a key NLP task where each word in a sentence is assigned a tag indicating its syntactic role. Prior language experience doesn't give an advantage. The task involves building a hidden Markov model that infers the most likely sequence of underlying states based on the observed data, regardless of the number of possible tags.