Key finding: "International Movies" is the single largest genre tag (2,752 titles), overtaking even "Dramas" (2,427). This tells us Netflix's catalog is heavily globalized. Comedies (1,674) rank third, making Comedy + Drama the dominant content combo. For our recommender, genre weighting matters — a movie tagged as "International" alone carries less discriminative power than "Thrillers" or "Sci-Fi & Fantasy".

Key finding: Netflix experienced explosive content growth from 2015–2018, peaking at 1,147 titles in 2018. Content dropped slightly in 2019–2020 (likely due to production delays from COVID-19), then dropped further in 2021 as the dataset captures only part of that year. For the recommender, newer titles (post-2015) dominate the corpus, so similarity scores will naturally skew toward recent content.

Key finding: TV-MA (adult/mature) is the most common rating at 3,207 titles (36.4%), followed by TV-14 at 2,160. This reveals Netflix caters primarily to adult audiences. For a personalized recommender, rating should be a filter option — family users need PG/TV-PG content, while mature audiences can explore the full catalog. Our system includes this as a live filter.

Key finding: Director has the most missing values (~2,634 rows), followed by cast (~825 rows) and country (~831 rows). For the TF-IDF vectorizer, we rely on description and listed_in (genres) — both have near-zero nulls, making them the ideal features for our content-based engine. Missing director/cast are replaced with empty strings before vectorization.