Natural Language Processing Roadmap – Part III of the NLP Roadmap

March 11, 2022

In the final part of our three-part series, we will see our roadmap advance from Deep Learning to a Natural Language Processing Roadmap.

There is no finish line, just another starting point

First of all, congratulations on coming this far in our journey Towards Natural Language Processing. In Part I and II of our journey, we covered our programming basics and then ventured all the way to Machine Learning and Deep Learning. Now that we have come this far, we can finally tackle our Natural Language Processing roadmap.

So without any further ado, let’s have fun with it!

The Natural Language Processing Roadmap

Ok, let’s start at the beginning. If I put the Deep Learning roadmap as a “prerequisite” of the Natural Language Processing roadmap, there is a reason. Frankly, taking the time to go into Deep Learning is going to help you so much. You will have such a better understanding of the processes going on under the hood of the libraries and packages you are using, and you will understand why you do what you do. So you will see that DL is part of this roadmap as well, but I would suggest you really treat it as a prerequisite.

Also, a tiny disclaimer: I will try to write an article for each of the subjects listed in this roadmap, but it is definitely going to take a bit of time, so please, bear with me and be patient. You definitely have already enough material to go through anyway 😉

A True Must: Natural Language Processing Specialization by DeepLearning.ai

Ok, let me spill the tea. This course is so good I took it twice. Once when I first started my PhD and I was trying to freshen up my academic understanding of NLP, and once a year later when the organizers updated the course to include explanations about Transformers (which I was actually craving at that point). You will cover most of the important subjects you will find on the Natural Language Processing roadmap, and the course will give you a good enough basis for you to cover the rest of the subjects on your own.

CS224N: Natural Language Processing with Deep Learning by Stanford

This course is also a superb resource. So good, in fact, that I can’t believe it is 100% free. When it comes to Natural Language Processing, Stanford University is one of the big names, if not the biggest name. This course is a series of college lectures, which means the videos are longer and dense, but frankly I think it is great. It really pushes you to pay attention and dissect the topic in order to advance.

Making an effort will really push you forward in your understanding of the field. In the lectures you will find a blend of theoretical concepts, practical issues, and suggestions on how to successfully complete a NLP project. The lecturers do an amazing job at explaining more complex concepts, and the course encompasses a variety of topics. Definitely a must if you enjoy college-style lectures and want to dive deep into Natural Language Processing.

In the title, I linked the 2019 series of lectures. In the 2021 series of lectures, a few topics have been skipped and a few added. I suggest you take a look at both and try to cover most topics by blending the two versions of the course.

Code-First Intro to Natural Language Processing by fast.ai

If you remember, in the Deep Learning roadmap, I strongly recommended following the fast.ai Practical Deep Learning for coders course. The same organization also provides this series of videos about NLP. Some videos are quite short, others are a bit longer, but all in all, this course as well provides you with good basis about most of the subjects I put in the Natural Language Processing roadmap. However, as the title suggests, this course is more code oriented and more hands on, which is great to help you along your first projects and to develop your first portfolio.

Neural Networks for NLP by Carnegie Mellon University

As the title says, this course is more Neural Networks oriented, but it goes into such depth about each topic, I think this is a true resource if you want to have a deep understanding of the subject. What’s even better, the course syllabus provides an incredible array of reading materials, great to really plunge yourself in NLP.

Deep Learning for Natural Language Processing by the University of Oxford and DeepMind

Similarly, the Oxford Natural Language Processing course offers a true in-depth look at NLP, and covers many of the subjects present on the Natural Language Processing Roadmap. This course is really well-rounded, and is not only wide but deep. The reading materials and the proposed projects allow you to explore the theoretical as well as the more practical aspects of the science.

A few more resources to Keep in Mind

Now, the course we discussed are all amazing and will give you a very good knowledge of the field. Each one will cover most if not all the topics in the Natural Language Processing roadmap, and most of them will also get you started on building your portfolio. But once you’ll be done with the studying, you will want to move on to real projects, and it is nice to have a little cheat-sheet handy.

So here are a few more resources and tutorials – as per usual, I will add to the list as I go, so make sure to check it out from time to time!

Advancing on the Natural Language Processing Roadmap

So there you have it! Congratulations on all your hard work, keep it up! And if you ever feel discouraged, don’t worry! That’s totally normal; all this stuff is hard! Take a break, go for a stroll in the park if you can, reach out to someone who has faced your same problems, and come back refreshed!

And as usual, see you in a bit!