Avis et commentaires pour d'étudiants pour Big Data Integration and Processing par Université de Californie à San Diego

Good course. But we couldn't relate the theoretical videos with hands-on.

Teaching is good but practice software provided completely outdated.

The final assessment somewhat not really well designed, imho.

Pyspark environment is quite not ready for this course.

spark hands on should have been more instructive.

mam teaches well while sir is not good at all

Good info, just a lot of info to digest.

All is good except the Splunk case

Very difficult final exam.

U​nmaintained resources.

Tough one.

The course and presentations were very informational and good. I enjoyed that aspect. I would have rated the course 4-5 star based on that. The reason for the low rating was the 6th module, and the fact that I felt very ill prepared for the syntax of creating all the operations in Spark (building out the Jupiter notebook). We really did not cover much of that, and it was quite punishing to search the web and sources to make things work. Even the instructions to export the csv file were misleading at best. I have a full time job and a family, I am not some young undergraduate with copious amounts of time to waste. While I am not opposed to some searching of other sources, I would like to have more of the useful information taught so that it is not so much of a burden. I believe that if you are going to test people on something, you should at least cover it in some sense.

With deep regrets I feel obliged to share a negative rating on the course. While the course material/video lectures are average to good (no rocket science but well done introduction into the subjects), the hands-on exercises and particularly the technical environment, i.e. Cloudera VM is entirely messed-up: - setup scripts are not working/ are outdated (e.g., anaconda requires no-check-certificate); user permissions are all set wrong and need to be corrected; firefox outdated with update function not working; countless error around spark context (SC) variables.... and so on... For a course that is so prominently promoted on the platform the least expectation is that the provided environment works and that students don´t need to spend hours on google to figure out how to debug the cloudera image.... Here, imo, a much better job can be done!

Although the contents of the course is good, I found that the hands-on exercises needed to pass tests were problematic due to many errors occurring when trying to setup the tools or running provided scripts. This means most of the limited time I have for this was spend browsing the course forums and the internet chasing solutions for errors occurring in the exercises and not on actually working on the assignments.

Also the course makes you install several tools/apps. In itself it is good to be exposed to these tools, however some of these are only used to a limited extent, while still taking time to install and setup. Worst is one of the tools requiring personal information in order to be downloaded and as a consequence being chased by sales reps for the tool.

Again I'm torn between quitting this specialization and biting through the rest of it. While the course is good on the high-level view, the link to the low level, the tools and their application just doesn't work well for me. The different tools are presented and used just enough to scrape a tiny lttle bit of the surface, then you're heading on the next chapter. Like in the previous courses, the tools in the VM sometimes need quite a bit on tinkering until everything works as expected. The main drawdown in this course was the final test that I did not felt prepared for at all. On the bright side, you learn to love the Spark manual...

PLEASE, UPDATE THIS COURSE AND BRING IT TO TODAYS LEVEL. IT'S ACTUALLY BETTER THAN THE AVERAGE FEELING IT LEAVES BEHIND.

Course materials are dense in high-level knowledge whereas the final project is technical. The hands-on learning is too linear and hand-wavy to leave users a programming assignment at the end of the course without the same constraints. The course was great in general for the subjects it covers. I do not think the applied hands-on learning is done in a way that gives any lasting understanding. I've given only two stars mostly for the reason that it took over 2 hours of my time simply debugging issues with the included VM to allow me to complete any course work. I would not expect that someone without a systems background would be able to complete the course work at all because of the several core issues with the VM configuration and included packages.

Overall this course does have some good content and delivers big data concepts. However, as others have mentioned some content (especially in early modules) could either be combined or ommited. Key focus areas on Spark and MongoDB are not given enough focus and lab time.

The quizes have badly worded questions. Finally the last assignment required to pass the class has bad directions and covers content not reviewed in the class. Spent a frustrating amount of time trying to get what most likely is simple code to work.

I'm hoping this particular module is revised. For those just interested in learning Spark or Mongo and not doing the certificate program you can probably get better learning from doing your own research.

I'm about halfway through this course and the specialization as a whole.

It it apparent that these courses were created a few years ago and have been left to their own devices since then. Any software version that you need to download is so old it may not even exist, and if you want help with it don't count on any responses.

Also as this specialization goes on, it seems to be more and more abstract, wordy lectures where you will absorb very little, and then a rushed assignment where you try to apply something literally one time before they move on to the next item.

With this level of engagement and assignments I will end up having to actually learn this stuff elsewhere from someone who knows how to teach.

I think that this course doesn't provide a substantial value to the student. It's basically a series of theoretical videos with irrelevant exercices that the student doesn't even have to think about. It's only about copy and paste until the last assignment. Until then, it's just a waste of time. Obviously it will be a good course for those people who only want the certificate and to pass the course with no effort at all, but it provides no value. On top of this, there is no technical support and I have struggled a lot in order to make everything work properly. I also suggest Coursera to give some guidance in the last assignment, there is a lot of lost people.

Again, another course in this series shows a lack of effort in its quiz construction. By the final week, you are presented with a challenge that will require numerous hours pouring over different documentations of both pyspark and MongoDB because there is a lack of essential knowledge being taught in the course. The final "project" is based on a very small amount of what was learned, and as it so happens, only a small amount of what was needed was actually taught. I'm hoping for improvement with the rest of the course, because the majority of this course was good, but the final week just ruined the experience.

The last assignment of the course is, compared to the others, more difficult. In my case, I ran into several errors which I couldn't get help in solving by using the course Forum, as the end of course deadline was just a few days ahead. I had to analyze the tweet texts for the last graded assignment without using Spark framework (nor any of the other "Big Data" tools explored in the course).

I also found some of the videos by PhD. Amarnath Gupta were difficult to understand, his examples were unclear and, in my opinion, too complex and difficult to follow and understand what was the reasoning thread.

I was really disappointed by this course, having found the previous courses interesting and helpful. I found the standard of teaching and explanation was poor, and difficult to follow, and the exercises, especially the Mongo DB and Spark remarkably difficult to work out with very little help or support given.

With Spark 2.0 and tools which run over Mongo to provide a SQL interface I'd challenge the usefulness of learning interfaces which are well out of date. We should now be learning SQL interfaces over both Mongo and Spark. The existing interfaces are difficult to just get in the way.

While there was a lot of useful information in this course at a "survey" level, the slides, documentation, and content in no way aligned with the level of knowledge needed to execute the assignments. That, coupled with the utter lack of maintenance on this course which resulted in MANY issues with incompatible tool versions and therefore code crashes, made this an extremely frustrating experience. If this had been my first Coursera experience it would have been my last as well.

IF YOU ARE GOING TO PROVIDE COURSES MAINTAIN THEM AND MONITOR THEM!

This was one of my favorite courses. i learned a lot of new things however the reason i am rating it low has to deal with the whole section on the hands-on; which in my opinion is the most important part. the instructions on downloading and running Anaconda was not clear and i had some many error messages that i was not able to do properly any of the assignments. besides no staff support at all. the professors are great and their explanation fantastic. should revisit the whole hands on and provide more staff support. a pity!

Most of the course content is similar to the rest of the specialization. Week 6, however, is a sudden spike in difficulty using mongoDB and Spark without any explanation of how to use the advanced capabilities of either language. The last two assignments were nearly impossible for me to do while googling how mongodb and pyspark work. Either would have been fine if the course focused on one or two tools in depth, but not with the survey style that it uses.