User's Facebook behavior was my source of inspiration.
What it does
Emotions are essential to effective communication between humans. Our target is to learn machines how to classify text as positive or negative. For example, by using this model, user's Facebook are able to recognize positive or negative post on their timeline and can improve their reaction to a post according to a context. (We will never find user liking a post where someone is died!)
How I built it
I built it in using the powerful pytorch library: Torchtext
Challenges I ran into
Gathering data is a big challenge. I have used a lot of datasets found on the web which are a valuable source of information.
Accomplishments that I'm proud of
I am pround to accomplish this great project. It was not easy but I finished it!
What I learned
I am able to train Machine Learning model which can analyze and classify text it as positive or negative.
What's next for Facebook Post Analyzer
I want to improve this project; there is a lot of things to do in the future!