While some of the nuances like how to build word embedding, intuition behind multi-dimension vector transformation logic are not clear or hard to grasp, but am able to understand and familiarize the concept of vector space models with k-nearest, linear regression and naive Bayes algorithms. Very good course material, will pursue the remaining courses and highly recommend to others.

Very nice programming exercises. The grader is quite strict regarding the choice of numpy functions: e.g. array.reshape(-1) is just as good as array.squeeze() but was counted as wrong; v * v is just as good as np.square(v) and was counted as wrong. As a statistician, I also found confusing that you talk about likelihoods but actually you mean likelihood ratios (Bayes factors).

Nice introduction to NLP course. Sometimes I felt that some parts were a bit disconnected (for example, in Week 4 we talk about word translation, but end up not using it). I felt that different pieces were glued together to create the course but some of then didn't match very well. But my overall opinion is that this course is good. You should try it!

Interesting course, and great lab coding. Learned new interesting simple approach to word embedding that will help me in my work. The only issue, as first time user of Coursera is trying to find help. It is not so obvious who to contact when things don’t work. I’m looking forward to going further and taking the Attention course

The presenters often looked above the camera, its like a weird thing you would notice. Any I am convinced that the answer for the Tehran/Turkey question in the practice quiz for week 3 is wrong. washington to US is a [-5 +1] change, and applying that to tehran, you get [4 2] which is closer to japan at [4 3] than turkey at [3 1]

I am not from a Computer Science background and I am not familiar with python programming. So, the course was a little bit tough for me and took me a considerable amount of time to finish. The teaching was up to the point and very precise. I mostly understood the lectures, and logic to solve the assignments were straightforward.

The course was overall great. I did learn a lot from the course . It gave me a new way of solving different problems like that of sentiment analysis. The last week was really great where the concepts of translation were taught. I think it could have been better if the concepts of locality-sensitive hashing been more elaborate.

Very basic stuff although good for beginners for building the foundation. Programming assignments can be improved further. One is expected to guess and use some pre-defined functions like np.vstack. If you fail to guess what the tutors want and implement something else, no marks are given although the function is correct.

The content and pedagogy of this course is excellent, but there is far too much hand-holding in the programming assignments. I have some doubt whether a person with a passing grade necessarily has mastered that content, or merely is able to carry out step-by-step instructions requiring only basic Python competence.

Overall great class.

It was odd that we're given word embeddings without much explanation for how they are created. I would have like to create embeddings.

The final code block of the fourth assignment could be refactored to two or more functions to make the assignment much simpler without losing meaning.

Buen curso con buenas explicaciones e interesante material práctico. Creo que a veces las explicaciones resultan un poco cortas y en los laboratorios uno está un poco sujeto a usar funciones de Python muy precisas que le restan algo de flexibilidad al momento de hacer la evaluación de la actividad

Very easy to understand and provides useful techniques I was able to apply immediately. Only downside is the exercises seem a bit on the side of spoon-feeding. The most extreme example being where there's a comment telling us to increment a number by 1, and we literally have to put a 1 there.

The content is really interresting but the programming assignments are not as challenging as they should be. I was expecting more in-depth programming of the mathematical concepts and what we were required to do was just to implement some basic python code around the interresting functions

Its well structured course, that builds the foundation in a very simplistic way in its videos and then gives practical implementation using python in exercise. Although there are lot more NLP concepts that are missing in it which I believe will get covered in the remaining specialization

It was a fairly good course, but I still expected a bit more from it. The coding assignments could have been improved. rather than making the assignment "fill in the blanks" type, make it more research project oriented. Overall a good exp, learned a few new skills. Thanks!

The course is very good in building intuition of important NLP topics (lectures especially). I would love the labs to include non-numpy tools, otherwise much of the effort goes into figuring numpy out, rather than trying to use the intuition from the lectures in practice.

The overall subjects are OK. I think this specialization could be 3 courses instead of 4. Some videos are too short and needs some more explanations, especially the LSH was unclear for me. The last assignment was harder than usual for me, especially the last part.

Props / videos which show real visualization can improve by a lot. Exercises though easy to medium, can be frustrating sometimes. The comments given for guidance are confusing sometimes. Support needs to improve rather than asking us to put on the bulletin board.

Quite good - no idea why the one trainer tells us in the videos he is going to teach us something and then the other trainer does all the work. And at the end he comes back to tell us what he taught us - but he did not. All he does is intros and outros of videos.

Although the course is great, the video lectures could go into more detail regarding how exactly to implement the theoretical concepts taught in python. The lack of this implementation explanation makes the weekly programming assignments unnecessarily difficult.

Good overview, it is quite basic especially if you have previous background on NLP but helps to you have an introductory background of the most basic techniques when treating word embeddings. Eager to keep the specialization to see how complex it would get!

There could have been more depth added to this course, this would make it more interesting. I liked this course, but on some topics, it seemed more of like we just touched upon and finished. So, adding more insights may be more useful to this course.

Very well thought out course, great connections between the weeks and everything builds up into real world applications. I am fairly certain there a small number of errors in some math and some of the hero videos would not load for me.

Great concept and great teaching. I know this gets harder to evaluate afterwards but the labs seem too guided. In the end, we are wondering if the concepts are coming from us or from the laboratory formatting.

Great to begin with NLP

I was enjoying this course. But last week was not good. In the last week's assignment you have given everything, and have no clue of understanding why when how. I am disappointed because of last week. It was complete waste of time.