I was very much looking forward to the course on text mining, but my enthusiasm has been dampened as the course was not designed particularly well. The hands-on examples are relatively short and stayed on a superficial level. They did not prepare students sufficiently for the assignments, which were pretty buggy and felt like a wild goose chase at times. On several occasions, I seriously considered breaking the honor code and looking up the right answers, as this would have given me the time and opportunity to actually learn something new on LDA and similar topics from online tutorials that are available on other sites in the internet. Having said that, I still learned something in this course, but what I learned clearly remained below my expectations.

I have been working through the entire specialization, Applied Data Science with Python. The first two courses of this specialization had a lot of attention to detail, the assignments were well laid out and challenging, and the addition resources linked by the instructor were really helpful. Moreover, the lectures themselves were more engaging and segmented.

This course was less informative than the other courses I've taken in this specialization. You would be much better off purchasing the O'Reilly Text Analysis in Python book and reading through it in more detail prior to taking this course or in-between lectures.

This course is much lower standard in comparison to the rest of Applied Data Science series. There are significant issues:

1. The concepts are covered either at very high level or completely missed

2. Assignments are of mechanical nature and lack the link with real life

3. Poorly written assignments - key info missing and unclear

4. Auto-grader is very picky and will require loads of ours of troubleshooting

I would not do this course unless you have to as part of the specialization.

If you do make sure to read discussions before attempting assignments- trust me it will save you a lot of trouble

This module of the Applied Data Science with Python is incredibly poor in comparison with the other modules. The assignments sometimes are not related to course material, there may be 1 practical/lab for every week in the course - NOT enough to help with the assignment. Most videos don't have accompanying notebooks to actually apply concepts, meaning that I will have finished this module with little skill improvement. Overall, in comparison to the other modules, I am severely let down by this course, too much theory not enough practical work.

After first 2 courses in this specialization this was a real disappointment. The course felt rushed and mostly dealt with how. It was real short on content related to application and why things are being done. Even the assignments didn't provide clarity on how the results are to be interpreted and what could be ther real world implications.

Material needs to be expanded to ensure that this course is not just to show how python can be used for text mining but also to include examples of where and how this is useful.

The overall material was good. That being said, this is the first time I have taken a MOOC course and felt like 90% of the time I spent was fighting with the auto grader. The instructions in many instances were unclear, so when you are dealing with a grading system that grades items as 100% correct, vs 100% incorrect with really no feedback as to what you did wrong it can be very frustrating. Without the Discussion forums there is no way I would have ever figured out what to do for some parts of assignments.

The course leaves you in the dark. Firstly, the assignments questions are written very poorly, making students waste their lot of time in understanding and finding what actually they need do it for auto grader to asses their solution. Secondly a lot of concepts are missing which are essential for the course and last but not the least there is no reasoning why are you doing some things in a particular way which is more relevant for online course which rather can be obtained just by reading some texts.

Out of the 5 courses in the specialization, this course was not up to the level of the other courses. Full of theory not much practical explanations and there wasn't much practice modules in each week just like other modules.

For Assignments, i was not even able to refer any module in this course to check for syntaxes. It was very tough for me to solve assignments as there was no reference in the videos or practice modules.

Need huge improvements in the course.

Course packed of information and topics in four weeks so it feels sometimes rushed.

Especially the forth week (topic modelling, information extraction, semantic similarity and generative models all in one week) feels disconnected from the rest .

The exercises do not help too much, with several mistakes and ambiguity.

Nevertheless, the theme is really interesting. Possibly the errors can be corrected in the next runs.

Plus for using Python and NLTK.

The course is very boring & way of explaining concepts is not great at all.I have wasted lot of my time in understanding the assignment questions & sorting the autograder issues .I guess its high time that teachers should revamp the course considering the quality vs the price a student is paying for the course .

There are better courses on udemy which I could have taken & they explain concepts in a very simple manner rather than boring methods.

Content in the course is interesting and given the amount of data stored in text very valuable. However, I would encourage the staff to provide more coding examples. I would also suggest moving away from assignments and towards projects - (a) projects would likely force more comprehension instead of code shopping and (b) the autograder is terrible: I can't believe the amount of time I wasted because the autograder was not set up properly.

I was disappointed by the lectures in this course. My impression is that extremely complex concepts are mentioned in passsing and poorly explained, while a large amount of time is spent on trivial examples. The programming assignments are more interesting and appropriately challenging (compared to other courses in the specialization), but leave me without any confidence that I could accomplish a text mining task in python independently.

It's really unacceptable that there should be errors with the autograder (which were left unfixed) and I wasted a lot of time trying to debug code which was actually working. As well this course did a good job with the introduction to the concepts in the first two weeks and then dropped the ball with content that appears rushed and disorganized. The LDA and other concepts need to be presented better.

First Coursera course I've taken with mistakes in the material and the grader. Also the first course where they mostly decided not to provide notebooks to review the material but instead made you scrub through the videos to find the actual code. Lastly the assignments were not terribly well tuned to the lectures (which were decent) and didn't make me feel like I had a great grasp of the material.

A lot of exercises have unclear instructions (see discussion forums). The exercise on topic modeling especially was a waste of time, you're not really learning anything by running these small pre-frabricated scripts. In general the exercises were extremely shallow and did not require any creativity or actual problem-solving, in contrast to some of the earlier courses in this Specialization.

One of the more disappointing classes in the U of M data science specialization, due mostly to inconsistent quality of the assignments. The videos are interesting but lacking in detail. The quizzes are trivial. Half of the assignments were OK but the other two were big time-wasters. The construction of this class seems just plain lazy. Proceed directly to google and skip this class.

This course has good content, but unfortunately suffers from major problems on the programming assignments. More time is spent digging through forums for help with incorrect grading from the automatic grader. Some assignments require that you submit the *WRONG* answer to get credit! It's a complete waste of time, which is unfortunate because otherwise the course would be great.

This course was not particularly well put together. I found the erratic behavior of the autograder for assignments to be a significant barrier to learning. This course was far more about battling data structures and python libraries than it was about text mining. The word "Applied" in the title should be replaced with "VERY VERY APPLIED..."

While there were parts of this course that were interesting, it is in bad need of an update. Most of the technical learing had to do with figuring out badly structured assignments and autograder issues. The criticisms of this course in other reviews are well-founded. This could be remedied by an update of the course by UMich.

Confusing explanations of NLP concepts. Inadequate explanations of how to use the Python packages to solve the assignment questions. I'm writing this review half-way through the Applied Social Networks Analysis course which is excellent and pitched just right. The contrast between the 2 courses couldn't be greater.

Los ejercicios de este curso tienen una dificultad muy superior a lo mostrado durante las clases, lo que hace que uno deba de invertir mucho tiempo en los mismos investigando en recursos internos. Por ejemplo, con una dedicatoria semanal de 10-15 horas me llevó 2 meses enteros hacer el curso.

Liked the simplified content. But minimalist approach w.r.to coverage of concepts - could be better. Tactical/Operational support, responsiveness from the TA w.r.to confusions on questions or grader can significantly improve. Thanks for the course, I learnt and enjoyed the hands-on sessions.

I am kind of disappointed of this course especially the lecturer were talking too much than showing the practical examples, for example 'topic modelling'. With a few slides of introduction about topic modelling showing some lines of code without any examples in notebook isn't helping a lot.

This is by far the weakest course in this certification and someone should invest in improving it. Very weak lecture videos, unclear assignments, terrible grader bot, lack of help in the forums, and an overall bad course. Avoid it if you don't need it for the certification.

Disappointing compared to the other courses of this specialisation. Some concepts were not defined (i.e bag of words) for the assignment. A lot of errors in auto graders, assignments. I had the impression to have a list of concepts and functions without a real explanation.