The course was well organized and the instructors are great.

Most fun came from the programming assignments.

The great thing about them is that there was no single "correct" solution for most tasks, students were free to explore different possibilities; solutions were then graded based on a data unknown to students.

Problem sets were quite boring in my opinion; many of them required calculations that are usually done by a machine; performing such calculations with pen and paper is quite error-prone (so for late problem sets I usually wrote small Python scripts; Excel spreadsheets would also fit the bill).

Programming assignments were quite time consuming (even for experienced programmers), but I feel that without completing these assignments the outcome would be much less.

This is a great course overall and I would recommend it to all people interested in Natural Language Processing.

Truly insightful and deep class. Very work-intensive, though.

If your interested in Natural Language Precessing this
is a good place to start. The workload is high so be warned. I was a little disappointed at the end, the syllabus said that they would cover question answering but they didn't show us a full semantic answering system but merely how to rip possible answers patterns out of the text.

I've taken this course a year ago. It was very interesting and challenging experience.

Pros.
- Interesting and broad course material
- Lecturers are active researchers in NLP area, and often provide information from recent papers
- very active and friendly forum ( BTW, this course was taken not only by people with no experience in NLP - I've met some people from Nuance on forum.)
- some programming assigments are very fun and challenging - you will never forget about PCFG's and Cocke -Younger- Kasami algorithm.
- I was very impressed when program which was written by me, performed better on POS tagging than I do.

Cons.
- some lectures and assigments are very boring
- very annoying errors in homeworks. Hope in next sessions they all was fixed.

Resume:
Recomennded for everyone, interested in NLP. But forget about 6 - 8 hours per week - this course will take much more time. But the result is worth the effort.

A difficult, but very rewarding course. The programming assignments test the techniques learnt on a weekly basis, and feel like real applications of the theory, rather than simple university assignments. Highly recommended.

Great class and the programming assignments are challenging and insightful. This is the best way to learn basics and applications of NLP. Highly recommended!

Very interesting class. This course covers many topics from Natural Language Processing & Information Retrieval. But comparing with Machine Learning course, this course requires that you spend more time teaching yourselves using the books from authors of this course (I had both books already on my bookshelf, so I had a chance to ready them at the end ;-) But beta-version of both books are freely accessible in Internet, so you don't need to buy them).
Some home works were very challenging (especially 4,5 & 6), but some were very easy.
Anyway, I'm happy that I had a chance to study this course directly from prominent scientists!

Never could imagine that task of processing a human-written text can have so clear and useful solutions. I highly recommend this course to all web developers. Very insightful!

This class was great, bettered only by andrew ng's ml class and cs212 from peter norvig. They got right into practical matters in nlp that helped me in kaggle's asap2 competition.