As young people we understand the struggle faced daily and how life can be hard sometimes. We all need a pick me up every now and then. So we just wanted to create something that makes the world little better for someone out there. #AskOnePositiveBot #SpreadPositivity
What it does
It a twitter bot that targets negative tweets and sends self-generated positive tweets to improve the users' mood. Use #AskOnePositiveBot to talk to the bot.
How we built it
We use natural language processing, especially sentimental analysis, to detect tweets with a negative sentiment and store the users. We use Markov chains trained on a big data set that includes a variety of inspirational quotes, books, jokes, positive tweets, etc to generate encouraging tweets for the above mentioned users. The bot is constantly learning and improving from its previous generated tweets.
Challenges we ran into
Obtainment of adequate training data; Determining the target users of our tweets. Training the bot appropriately to generate coherent responses.
Accomplishments that we're proud of
Creating something that is actually useful and might make a positive change in someone's day.
What we learned
The more we train the bot on the data obtained from twitter the less coherent and irrelevant it becomes. Python is the best language we have used! The Internet is filled with all kinds of emotions and sometimes a little bot can do a lot to inspire by creating a supportive feeling.
What's next for One Positive Bot
Improve its training data quality. Target a more specific category of users. Implement memes for the lols.
Log in or sign up for Devpost to join the conversation.