Inspiration
"Words have energy and power with the ability to help, to heal, to hinder, to hurt, to harm, to humiliate, and to humble." -Yehuda Berg.
The idea of this project is to understand who a person is before following them in order to decide quickly if they have content that we would like to see over time. Imagine if we could understand people before following them online, if we could predict whether or not their words will hurt us or make us upset. If we could see whether a persons' words are harmful or helpful, we could be better informed about who we follow and what content we allow ourselves to be surrounded by. This application can help people to live more positive lives by surrounding themselves with more positive, healing, and helping words.
What it does
You can enter the persons' Twitter handle and see if they are posting positive or negative content using sentiment analysis. We want to be able to analyze tweets and see the connotation of what they are saying. We want to use machine learning to eventually be able to output pie charts of how often people post positive, neutral, and negative tweets. Inspired in part by machine learning word clouds, we would like to see what words people use the most.
How we built it
1) we used twitter API to fetch data from twitter for a specific user 2) we pre-processed our data to a format which is appropriate for our machine learning model i) converting text data into numeric representation using TFIDF 3) we used natural language processing techniques i) bag of words for sentiment analysis ii) word cloud 4) we used bootstrap, js, html, css for our frontend part ( credit to : https://github.com/StartBootstrap/startbootstrap-sb-admin-2) 5) we used flask as a back-end mostly we used it to write an API
Challenges we ran into
1) twitter API data access limit since machine learning model require too much data we can not get tweets as we want to because of twitter API's limit. 2) short amount of time to train a neural network model for our text analysis.
Accomplishments that we're proud of
- we built something which can change the way we interact on the internet forever with in the weekend.
- The idea can turn into a bigger project not only it can be used on twitter, but also it will be used on any other social media. ## What we learned
- Many great things in the world created by those who bother the current working of things, so we try to change the status quo by first understand then follow than vise versa. ## What's next for Psychic computing machine
The future goal is to use Tweets to understand what kind of person someone is. We want people to be able to have such key insights without having to pore through countless Tweets from the users they want to know about. We would love to see this application become integrated into the Twitter universe.
Log in or sign up for Devpost to join the conversation.