MovieLens SQL Rating Explorer
An interactive SQL analysis of MovieLens user ratings by Shalynne Orth.
Project Overview
This project uses SQLite, SQL, Python, and a reactive Marimo app to explore MovieLens movie ratings.
The project asks:
Which movies have the highest average user ratings when they have enough ratings to make the result meaningful?
The app joins the movies.csv and ratings.csv tables using movieId.
Users can choose a minimum number of ratings with a slider, then view the ten
highest-rated qualifying movies in a table and bar chart.
Key Results
With a minimum of 50 ratings, The Shawshank Redemption (1994) was the highest-rated qualifying movie, with an average rating of 4.43 out of 5 from 317 ratings.
| Movie | Rating count | Average rating |
|---|---|---|
| The Shawshank Redemption (1994) | 317 | 4.43 |
| The Godfather (1972) | 192 | 4.29 |
| Fight Club (1999) | 218 | 4.27 |
Analyst Insight
Increasing the minimum rating count from 50 to 100 removed movies that had fewer than 100 ratings from the top-ten ranking. The revised list emphasized movies supported by more audience feedback.
This demonstrates that the rating-count threshold changes the amount of evidence required for a movie to appear in the results. A higher threshold makes the ranking less influenced by a small number of ratings, although it does not make the results an objective measure of movie quality.
Run the Interactive App
From the project root folder, run:
```shell uv sync uv run marimo run src/datafun/movies_notebook.py