A 2-page interactive dashboard analyzing Netflix titles across genres, countries, ratings, and content type — built on a normalized star-schema data model with 6 related tables.
| Page | Description |
|---|---|
 |
Overview page — trends, genres, map, ratings |
 |
Detail page — cards, pivot tables, country map |
Before loading into Power BI, the raw wide-format CSV tables were normalized in MySQL using UNION-based queries to unpivot multi-value columns into clean long-format tables.
| File | What it does |
|---|---|
sql/create_listed_in_table.sql |
Unpivots 3 genre columns → one row per genre |
sql/create_directors_table.sql |
Unpivots 13 director columns → one row per director |
sql/create_countries_table.sql |
Unpivots 12 country columns → one row per country |
sql/create_cast_table.sql |
Unpivots 50 cast columns → one row per cast member |
See the
sql/folder for all query files.
Normalized star-schema: netflix info (fact) linked to category, country_released, cast, director via show_id
The raw Netflix dataset was normalized into 6 related tables using Power Query:
| Table | File | Key columns | Used in visuals |
|---|---|---|---|
netflix info |
info.csv | show_id, type, title, date_added, release_year, rating, duration | All pages |
netflix category |
listed_in.csv | show_id, listed_in_1/2/3 | Top 10 Genres chart |
netflix country_released |
country.csv | show_id, country_1 ... country_12 | Azure Map |
netflix cast |
cast.csv | show_id, cast_1 ... cast_50 | Cast pivot table |
netflix director |
director.csv | show_id, director_1 ... director_13 | Director pivot table |
netflix description |
description.csv | show_id, description | Page 2 — description card (displays show/movie synopsis on selection) |
All tables linked back to
netflix infovia show_id
| Visual | Type | Fields |
|---|---|---|
| Shows added by Date | Area chart | date_added (year hierarchy), COUNT(show_id), type |
| Top 10 Genres | Bar chart | category, COUNT(show_id) |
| Countries' Shows count | Azure Map | country, COUNT(show_id) |
| Shows by Rating | Column chart | rating, COUNT(show_id), type (Movie/TV Show) |
| Visual | Type | Fields |
|---|---|---|
| Countries' Shows count | Azure Map | country, COUNT(show_id) |
| Rating | Card | rating |
| Release Year | Card | release_year |
| Description | Card | description (updates dynamically based on selected title) |
| Movie/TV Show | Slicer | title (filtered by type) |
| Genre table | Pivot table | listed_in |
| Director table | Pivot table | director |
| Cast table | Pivot table | cast |
- Loaded raw Netflix CSV — 6,000+ rows with multi-value columns (genres, countries, cast as comma-separated strings)
- Split multi-value columns — used
Text.SplitandTable.ExpandListColumnto unpivot genres, countries, and cast into separate rows - Created normalized tables — each entity (cast, director, category, country) became its own query/table
- Handled missing values — replaced nulls in
director,cast,country, anddate_added - Standardized date formats — parsed
date_addedinto proper Date type and extracted year hierarchy - Built relationships — linked all tables back to
netflix infovia show_id
- Content volume grew significantly year-over-year, with peak additions in the 2018–2020 period
- Movies outnumber TV Shows across the catalogue
- International Drama, Comedies, and Documentaries are the top genres
- The US, India, and UK dominate content origin
- TV-MA and TV-14 are the most common ratings, covering the majority of titles