I'm super inexperienced with natural language processing, but a very common feature of NLP libraries seems to be part of speech (POS) tagging. In documentation, POS tagging is often referred to as part of an NLP pipeline. Which algorithms or techniques would take the output of a POS tagger as their input?