I'm looking for resources (books/articles/whatever) that provide mathematical formalization of NLP and statistical language theory. By that I mean clear exposition of the subject in terms of probability spaces (measure spaces) and so on. For example, many NLP books (like the Manning's one) use n-gram models which, as I see, may be modelled as Markov processes with word-states, but neither book states explicitly how the probability space for the process is constructed (I guess, there's something related to probabilities on formal languages?). I need such clear expositions. Thanks in advance.
2 Answers
$\begingroup$
$\endgroup$
My personal recommendation would be Introduction to Natural Language Processing by Jacob Eisenstein.
In this book you should find sufficient mathematical formalization/rigor. This books is also, in my opinion, a touchstone of many introductory NLP books.
$\begingroup$
$\endgroup$
2
I would highly recommend "Speech and Language Processing" by Jurafsky and Martin. It has everything with the mathematical and statistical basis that you need for NLP.
-
$\begingroup$ Actually, I went through this book and find it less rigorous than the Manning's one. In any case it doesn't contain what I'm looking for. Take, for example, chapter 3 devoted to ngram models. The author doesn't define probability space before diving into random processes representing sentences. $\endgroup$ Oct 9, 2022 at 17:25
-
$\begingroup$ Cool. I think that book assumes some of the fundamental things to be known beforehand. In that case, you may have to hop on a few other books in addition. I have not read the one recommended by @Ethan. Seems that one may help as per his comment. $\endgroup$ Oct 10, 2022 at 7:58