Very good work. For race you could start with the main characters? It would also be interesting to add demographic benchmarks on US/international cinema attendance. You can find this for the US on the MPAA site. Look for their most recent annual report. The UK has the BFI Statistical Yearbook. This would be a good way of introducing indexing and making a call on whether content is representative of cinemagoers. It's actually around 50/50. You could also incorporate different age ranges for characters and compare that with audience data. If you could get your hands on demographic data for each film in question, which companies do provide a a price, that would be even better. You could then perform the above exercise on trailers; to see if the distributor positioned the film differently during the marketing campaign in order to broaden its demographic appeal; to see if they changed positioning of the over the course of the marketing campaign (from teaser to feature trailer 2); to see if the positioning of the trailer had an effect on the final cinema audience. This would no doubt of interest to the likes of Women in Animation and other such groups by the way. There's a lot of this work going on right now in film and television. I think WIA would be really impressed. In the UK there's an ongoing data project called Diamond, a collaboration between the big broadcasters that aims to record demographic data for TV shows and monitor progress made in increasing the diversity of on and offscreen talent. That data may be made public once up and running. May be worth looking into in the future.

Awesome post! I'd guess that even if there were remarkably few female characters, the variety of words used by them in Rogue One would stand out, and this could serve as a sort of proxy measurement of character development?

SO sad, but I can't access the page with the vizualization!

As an emerging data scientist, I found this extremely helpful to read through your process and follow along with your code. And your writing style is quite personable, leading us through your narrative. And bravo for highlighting the issue of gender parity in media and culture!