Chandrasekhar Gudipati
7 Feb 2021
We data scientists have a knack for creating quick and robust solutions but fail to present them into
articles of information. This way, we serve no one but the select few data scientists
We're here to use NLP to get the necessary output.I could as usual put up the code but that would void the point of the previous slide. I will instead try and explain the algorithm as simply as possible in the next few slides.
Before we start, it's best to have a quick read about NLP. Here's an article that has immensely helped me: https://towardsdatascience.com/your-guide-to-natural-language-processing-nlp-48ea2511f6e1
In order to understand the meaning of phrases together instead of just words, like saying “not quite” is the opposite of “quite”, we split the sentences into groups of 2 or 3 or whatever (this number being 'n').
Through this, we can perform a rudimentary version of sentiment analysis with the given dataset and then come up with a model using NLP package that can predict the output given a part of a sentence
It's important to create a solution to this problem because literally every interactive AI application runs on it. Chatbots, Sentence fillers, sentiment analyzers and AI document scanners (OCR) are just a few examples of tools that woul benefit from our project
I am delighted to present the final version of my app on Shiny! It's the other link in the assignment button. In the Server part, I took the user input and used the NLP algorithm to get the output. I somehow wasn't able to generate the output reactively so you will have to refresh the page to get the value. Sorry about that :(
Thank you and have a great day!