How to learn NLP by practicing? Any learning path for this?


#1

Hi,
I have just finished with a basic course in NLP using python NLTK packages.It included projects on sentimental Analysis, Text Classification.

So, Wud like to know the path to build from this basic understanding.

What is the project area to start with NLTK package (Considering real time usage)?
Can learn other useful packages for NLP?If so, what are they?

Please provide ur insights.


#2

Actually, I am also searching for the same so if someone working in NLP can provide links for some MOOCs or a proper learning path for beginners would be great.


#3

Since you have the basics down, I would suggest you start working on real world data and learn by doing.

If you want to work on the code, considering that you know python, you can start working on analysing tweets. Twitter provides a fast and easy to use api that you can use to get real time tweets for you project. There are various real world text datasets on Kaggle, which you can use to continue developing your skills.
As for packages, check out Spacy and Gensim.
I would also suggest learning R, as it has advanced analytics algorithms for data analysis. The “tm” package for text analytics, will be a useful tool in your arsenal. “RTextTools” is another great package for Text Clustering and classification.
I personally use python for all my data parsing needs, and use R for analysis.

If you want to know how to go about analyzing text without worrying much about code, I would suggest you head over to Analysing sentiments in Movie Reviews. You can further explore the platform, ATH. They have introductory courses on text analytics, where the main focus is on “learning by doing”. If taking a course is not your style, you can upload your own dataset and start analysing it, from pre-processing to classification.


#4

Thank you, @Saurabh0424.