If you are interested in Natural Language Processing and want to use it to analyze text or a large number of documents, extract specific information, or even build software, but your programming skills are poor, chances are you’ll be overwhelmed by the content available online and for free. And this may lead to actually giving up even before jumping into it.

There are so many recommendations, so many courses, books, videos, and articles, but in the end, you don’t know where to start. I haven’t figured it out either, but I will be updating this post whenever I find something that can better pave the way. Here are a few steps that could help you:

1. Get familiar with the technology

Before starting to learn NLP, you should get yourself familiar with Python. You don’t need to become advanced in Python, it will be enough if you understand the basics of how it works, understand the most important concepts and do a few exercises. Practicing, although tedious, is essential for learning to code.

The best way to learn these things is to go through a course on Udemy or even on Youtube. You will need an environment in which you will be working, if you want to install it on your machine, just install Anaconda, or you can also work online on Google Colab.  

Get yourself familiar with Stack Overflow, which is a platform where you can ask anything about coding or check if anyone had already asked the same question before, and Open Ai’s Chat-GPT.

Time spent: I spent 1 month on the course, haven’t done anything afterward, so I forgot a lot of things, then I watched some other courses on Youtube at a double speed, just to remind myself and tried to do as many exercises I could. All in all, I spent 3-4 months of irregular studying and got to the point where I understand concepts, but I would need to look at the code in order to write something new. I noticed that I was better when I was practicing it for a few days in a row, so I highly recommend practicing it even for 10-15 minutes a day, as long as it is every day.

2. Find what interests you in NLP

The next step you should take is learning more about NLP. NLP has so many applications and so many things that you could learn, but it would be best if you start from the one that interests you the most. The easiest way to choose what to study would be to skim through a few books that have a good overview. You don’t need to read them, just go through the topics these books are analyzing and assess what would be good to be your starting point. It would be good if you can find something that is both easy, i.e., that beginners can start from there and also interesting enough for you to keep you motivated.

Top choices for books would be:

1. Natural language processing with Python. Steven Bird, Ewan Klein, Edward Loper, published by O’Reilly and available online: https://www.nltk.org/book/  (this one might be easier for complete beginners)

2. An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition, by Daniel Jurafsky and James H. Martin, published by Prentice Hall, available online: https://web.stanford.edu/~jurafsky/slp3/ed3book.pdf

Time spent: You shouldn’t waste too much time on this or go into analyzing and never-ending reading and researching without getting your hands a little dirty with the code. This is something that should be finished in a day or two, or even in a week, depending on how much time you have.

3. Practice, practice, practice

Once you have an overview of the field and know what there is to be studied and practiced you can choose one of the two approaches (or both) that suits you best when it comes to studying.

As you will see in these books, they cover a wide range of topics and are also practical, you can simply study along or you can find an online course, there are a lot of courses on Udemy, Coursera, EdX, Future Learn or some other similar platform.

The other approach is to find tutorials on Youtube or various blogs, like Analytics Vidhya, Towards Data Science, Towards AI, Level Up Coding, etc., and follow along. Be cautious not to just blindly copy-paste code, but think about it and experiment a little bit – it should be a study session.

Whichever approach you take, be prepared to get frustrated, there will be times when you will want to throw your computer out of the window, but this is how the study process looks like when you are learning to code as an adult. It will also take time. These technologies are developing every day at a rapid speed, so be flexible.

**IMPORTANT ADVICE: In order to learn a little bit faster, it would be good if you can think of some real problem that you can solve by using Python, or think of some process that can be automated. When you have a problem that is relevant to you, you will be motivated to look for the solution and on this journey, you will be encountering obstacles and finding ways to overcome them, and in the end you will see how much you have learned in the process.