Dataset
Finally, the grand finale is upon us, the last dashboard of the week and thankfully it was data from a source that was immediately recognizable to all movie buffs! myself included the IMDb database. This dataset was primarily based on IMDb’s non-commercial information on all movies, shows and short films etc. So, no box-office grosses and profits to revel over, instead it focused on the titles, artists, ratings and crews. It was a massive dataset and once more we were forewarned that prior planning and thought into what would like to uncover would save us a great deal of headache later.
Preparing the Data
Heeding this advice, I decided to harken back to my roots in India. I wanted to see how IMDb represents the Indian region over the years. India is one of the largest populations in the world and surely there would be certain nuances that are uncovered give the vast diversity of the region, people, languages and taste. With that in mind I created this simple workflow that first filtered for only movies that are released in India that are present on IMDb.
I used the following tables
title.akas.tsv.gz (primary table for titles and to filter for region)
title.basics.tsv.gz (identified the medium e.g. movies, runtime of titles and genres)
title.principals.tsv.gz (the role and category of primary casts in the film)
title.ratings.tsv.gz (ratings for the movies)
name.basics.tsv.gz (details of actors/directors and their names)
Once I had a curated dataset, it was of a much more manageable size to load and manipulate in tableau.
To prep my data for Tableau, I used Alteryx to stitch everything together from the massive IMDb files.
Data Exploration
Once I loaded the data into tableau, I began quizzing the data to see how movies are being represented on IMDb back in my home country. To be honest, I was shocked by two things, one the biggest movie database on the internet does not in its current state have the full range of moves being churned out back home, it’s missing a few languages and definitely missing a few big hits. Secondly, because of the community nature of IMDb, some of the titles were rubbish entered by fans of artists. Keeping this in mind I cleaned the dataset as best as possible and decided to present my analysis in four parts.
Analysis
Part 1: What languages are most films from the Indian region released in.
Part 2: How has the representation of artists changed over time? Such as Actors, actress and directors and who are the biggest stars based on Indian region.
Part 3: How have the top genres evolved with time over the decades based on ratings and their runtimes and votes.
Part 4: Does the runtime of movies impact their ratings? Are shorter movies more favoured than longer epics.
To answer all these and more click through to explore the movie dashboard and find out for yourself!
