MovieMatchMaker

Inspiration

Matchmaking movies to stars is quite a captivating subject, with all the star power and whatnot

What it does

The algorithm ranks 20 stars for an old movie based on % error of actual revenue; the web app then displays visualized data calculated by the algorithm.

How we built it

Scikit Learn (Python) for Objective 1; Oracle SQL database (currently unlinked but planned) for easier retrieval, React App (Javascript) for Objective 2.

Challenges we ran into

• Time constraints (after all, it's a hackathon)
• Definitely learning about new APIs/technologies
• Integrating project parts together

Accomplishments that we're proud of

• We got the web app up and running
• Our data collection/storage pipeline is fairly streamlined
• We implemented the RandomForestRegressor to calculate projected revenue given a star and a movie

What we learned

Lots!

What's next for MovieMatchMaker

Since this was a quickly-constructed functional prototype, there's quite a few things that we would like to improve on this. Primarily, we would like to expand on the various components included within our project and increase the cohesion between the separate parts. Additionally, additional research could be performed to improve the efficacy of our project.

Built With

• java
• javascript
• oracle
• python
• react
• scikit-learn