Computational Linguistics 4, Exercise 2: the lookup approach

Add "tables": "table" to the lemma dictionary. Then:

lemma.get("tables", "tables")  ->  "table"
lemma.get("chairs", "chairs")  ->  "chairs"   (no entry, so it comes back unchanged)

(Run with py -X utf8: the output was "table chairs".)

What it shows: the Stemming article says lookup tables are quick and handle
irregular forms well, but every inflected form must be listed, so new or
regular words can be missed. Each word needs its own entry, however
regular its ending. The Lemmatization article adds that a trivial way to
lemmatize is simple dictionary lookup, and that languages with long compound
words need rule-based systems, which can be hand-crafted or learned from
annotated corpora. That is why real systems combine a list for the irregular
cases with rules for the regular ones (the hybrid approach in the Stemming
article).