Computational Linguistics 1, Exercise 3: sorting tasks, and Georgetown

Using the NLP article's grouping:
- stemming: morphology
- parsing: syntax
- tokenization: text and speech
- word-sense disambiguation: semantics
- machine translation: applications

Georgetown as a warning: the NLP article says the 1954 experiment translated
more than sixty Russian sentences into English, and its authors predicted
machine translation would be solved within a few years; progress was much
slower. The Computational linguistics article adds that researchers expected
rule-based methods to capture lexicon, morphology, syntax and semantics, and
those approaches failed. A demonstration on a small set of sentences did not
show the method would scale, which is the lesson of Chapter 9 (a good result
on a small test is not a general result). That last connection is my own
reading, not stated in either article.

Source: Wikipedia, Natural language processing; Computational linguistics.