The reason why I am learning NLP is that state-of-art tools are able to recognize the key words to build chatbot or even data analysis for analysing the reviews from customers to improve services. Those useful applications attract me to investigate more about that.

Last week, I took a course for NLP in Kaggle. It provided step by step description and exercise to let you familiarize with the flow of NLP such as building models and training the data to obtain an accurate result. However, the tutorial gives a short brief of explanation which is insufficient to understand the mechanism and library of NLP. Thus, I found a YouTube Video to acquire more knowledge like how the chatbot reply to your answer correctly. It demonstrates a json file that contains the pattern words and the related response that showed me a clear concept about that. Here is the link of the video:

Bags of words Concept: https://www.youtube.com/watch?v=wypVcNIH6D4

Today, I enrolled in a course in Udemy to get some ideas of Bag-of-words that I have seen in Kaggle before. In short, Bag-of-words is a vector to summarize the quantity of words. The vector size can be defined by yourself but the course teaches the length should be around the number of words in the dictionary if you create a chatbot. Hence, you can enlarge the sample size in your dataset.