CS-431: Questions about the coming quiz and last year's quiz 4

Hi,

As there are no questions covering the Modern Neural Networks approaches to NLP in either last year's quiz or in the exercice sheets, I am wondering whether there will be questions concerning those slides in tomorrow's quiz.
Also, I have a hard time understanding the answer to the question 2.2 from last year's quiz 4 :
How can a lexicon associating roots and PoS to surface forms and a part of speech tagger explain the absence of any token related to a word ?
I would have guessed it was only a stop word list.

Thank you in advance,

Kenyu Kobayashi

Re: Questions about the coming quiz and last year's quiz 4

by Jean-Cédric Chappelier - Tuesday, 17 December 2019, 5:14 PM

yes, as I said in the lecture (and in the convocation), this new course can also part of the quiz.
??
it's not the resource that explains the absence: the resource is REQUIRED to explain their absence, in the sense that some module used these resource(s) to remove the considered token.
Stop list cannot be used here as, by the assumptions stated in the question, "can" is in fact ambiguous (a can of beer) and thus a stop-list cannot be used here (missing the PoS tag, useful to disambiguate can/V from can/N)