This project explores the profitability, ROI, and user engagement of various genres of film to recommend films for Microsoft to grow their new streaming service.
- Which genre got most profitability, ROI
- which genre got most engagement and ratings to get Microsoft movie good reputation
- which is the best season for movie release
- We found that Adventure, Sci-Fi, Animation, Action, Comedy, and Drama (in that order) were relatively high-performing on most or all metrics.
- Biography films were intersting in that they had very high ratings, although they performed poorly financially.
- Films with multiple genres also performed better on both financial measures.
we explored these datasets:
- Box Office Mojo
- IMDB
- TheMovieDB
- The Numbers
- Rotten Tomatoes
- List of films based on video games
- Film Adaptations of Video Games
-
Jeffsploration contains Jeff's Jupyter Notebook Jeffsploration uses imdb data together with budget data from the-numbers.com to create visualizations comparing genres as well as number of genres for films. Data wrangling and exploration is done with pandas. Genres are one-hot encoded so that a mask can be created. This allows values for each genre to be aggregated, and these aggregates include the genres when a film has multiples. Data visualizations are done with matplotlib.
-
hy_exploration_1: han's early EDA
-
hy_exploration_2: han's second EDA and plots exploration
-
hy_exploration_3: han's cleaned-up and Final Notebook. it covers data import, data cleaning, brief observation and plots.
-
Growing a Streaming Service.pptx: presentation doc
-
Plots folder contains exported images of seaborn and matplotlib images