As a former film student, I felt that my dreams of working in the film industry had come to an end after switching career paths to Data Science; however, that couldn’t be further from the truth. Data Science is now becoming commonplace in film, and there are four main areas in which Data Science is revolutionizing the industry:
- Film Production
- Marketing
- Streaming Services
- Scripting
Film Production
It sure seems like there are a lot of superhero movies being made these days, right? Data Science has a large part to play in that. Analytics is used in every stage of the movie production process, from predicting budgets to even choosing what movies to make or which actor should star in a movie. Warner Bros has even partnered with software company Cinelytic to gain insights into what movies to make in the future. While this provides fantastic insight for movie studios, it is partially responsible for why we are seeing more superhero movies, reboots, and sequels, as these types of movies are proven to be high performers for studios.
Marketing
Most big-budget movies are created with marketing in mind from the project’s inception. At the end of the day, movie studios want to make money, and Data Science drives a lot of these decisions. Studios will create targeted market campaigns by generating “micro-segments” of people. So, for example, a movie studio may find a cluster of Institute for Advanced Analytics (IAA) students who love to watch sci-fi movies about Artificial Intelligence, and they can target us appropriately with movie trailers about that particular interest. This generates a win-win scenario where people are suggested movies that they would like to watch, which in turn will generate money for the movie studios.
Streaming Services
Perhaps the most well-known use of Data Science in the film industry is with streaming services such as Netflix, Hulu, and Disney Plus. Data Scientists build recommendation systems to help suggest new shows to watch based on a user’s watch history, movie ratings, what time of day the user typically watches movies, and many more features. The recommendation system is perhaps the most critical piece of technology for these streaming sites in driving views, with around 80% of Netflix’s total stream time coming from recommended TV shows and movies. Recommendation systems are also popular side projects at the IAA! Here is a project completed by Class of 2020 students Cathy Tran and Jiin Son.
Another technique commonly used on streaming services is A/B testing. Have you ever noticed that the cover art for a movie on Netflix often changes? This is because Netflix is performing A/B testing with the cover art to see which image makes a user more likely to watch that movie or show.
Scripting and Simulations
While not common among big-budget films, many small independent filmmakers have been experimenting with AI-generated scripts, whether for a small stanza of dialogue or an entire movie. Deepstory.ai is one of the major players in this field, and generating a script is as easy as signing up for an account on the website and setting a few basic model parameters.
Artificial intelligence has also been used to artificially de-age actors (such as Robert DeNiro in the Irishman) and even to artificially generate audio based on a person’s voice. A recent example of this can be seen in the documentary, “Roadrunner,” which was about the life of the late Chef Anthony Bourdain. In the movie, a piece of dialogue was generated using AI to sound like Anthony Bourdain’s voice, as he had passed away prior to the film’s completion. While this was an incredible technical feat, its use has proved to be divisive and serves as a reminder that while this technology is exciting, there are still ethical implications to consider.
Data Science in the film industry is still relatively new, but its applications grow by the year. If you’re interested in finding out more about data analytics and film, I highly recommend browsing the Netflix Research website where there are some fantastic blogs going into detail about how data analytics and machine learning are used at Netflix.
Columnist: Gordon Hendry