What inspired us
- Social media has a lot of conversation going and it is rare that we find anything on twitter that directly tells us an actionable way to get involved in that issue
- Social media raises awareness and we wanted to use that awareness and make it more socially beneficial
What we learned
- We learned that it is not easy to analyze social media data and filter it out to create something that will find tweets that could help raise awareness.
- We learned more about natural language processing and how it's useful in a project like this
- We also learned about creating twitter bots.
How we built it
- Gathers twitter trends and uses natural language processing(via the vader package) to see peoples feeling towards a trend
- Naively isolates keywords from tweets for retrieving charity data
- Naively searches for charities and other information related to trend
- A twitter bot provides people with links previously gathered
- The Flask app does some visualization and displays bots tweets
Challenges
- There won't always be trending topics that require attention so thats why we have a test case text file
- The search for charities is still naive and needs to be improved before completely deploying the bot
- Tweets in foreign languages causes gaps in data
- We use limited data to make it usable for the hackathon, more optimization and more data will be needed for later.
- The bot was not completely finished as we wanted to make it reply to popular tweets
Log in or sign up for Devpost to join the conversation.