Natural Language Processing (NLP) involves using machine learning models to work with text and language. The goal of NLP is to teach machines to understand spoken and written words. For example, when you dictate something into your iPhone or Android device and it converts your speech to text, that's an NLP algorithm at work.
You can also use NLP to analyze a text review and predict whether it's positive or negative. NLP can categorize articles or determine the genre of a book. It can even be used to create machine translators or speech recognition systems. In these cases, classification algorithms help identify the language. Most NLP algorithms are classification models, including Logistic Regression, Naive Bayes, CART (a decision tree model), Maximum Entropy (also related to decision trees), and Hidden Markov Models (based on Markov processes).
Small insight before starting: On the left of the Venn diagram, we have green representing NLP. On the right, we have blue representing DL. In the intersection, we have DNLP. There's a subsection of DNLP called Seq2Seq. Sequence to sequence is currently the most cutting-edge and powerful model for NLP. However, we won't discuss seq2seq in this blog. We will be covering basically the bag-of-words classification.
In this part, you will understand and learn how to:
- Clean text to prepare it for machine learning models.
- Create a Bag of Words model.
- Apply machine learning models to this Bag of Words model. Here’s what we will be focusing on. Note: We will not discuss Seq2Seq, chatbots, or deep NLP. The materials I have used are from NLP with DL, so we will exclude the DL part.
To read the full blog: ML Chapter 7: Natural Language Processing
Top comments (0)