Inspiration
We wanted to compare products but we did not know how make an informed decision. There are alot of conflicting opinions online so its difficult to get an overall picture.
What it does
The user queries for topics that they want to compare .Our product uses twitter api to find twitter posts related to those queries. It then sends those posts to cohere, where our fined tuned ML model classifies the sentiment of the posts on a scale from 0-2. Finally the user can see the relative public sentiment of their search topics displayed in a Pie chart as well as a small sample of posts related to the queries. Users can use this visual representation to aid their decision making and gain insight on overall public sentiment.
How we built it
-On the frontend we used React, Material UI, and ChartJS. -On the backend we use Python, Flask, Cohere, SpaCy, MongoDB, Pandas, Twitter API
Challenges we ran into
- API issues, libraries broken
- Twitter data contains lots of advertisements, which pollute our data.
Accomplishments that we're proud of
- Fine-tuned Machine learning model using cohere
- Full integration with frontend and backend
- Created a tool that helps us make well informed comparisons
- Gain insight into overall public sentiment on topics
What we learned
- cohere api,
- machine learning
- frontend development
- training machine learning model
What's next for CApplePie
- More refined search (ex location, demographics, ...)
- extend supports for more comparisons. potentially stock markets
- scrape results not just from twitter but all of the web
- better machine learning model that classify posts more accurately
Log in or sign up for Devpost to join the conversation.