This post has been cross-posted from the SpringerOpen blog.

In order to determine the importance of a movie in the history of cinema, we usually analyze its success in terms of economic income, that is the amount of money it raised from tickets sold in theaters (the so-called “box office”), or in terms of critical acclaims by means of artistic judgments delivered by professional critics that evaluate a movie from aesthetic and technical points of view.

By referencing other movies, crew members show their appreciation and admiration, and drop hidden messages for attentive audience-members.

Recently, researchers in the field of social network analysis have proposed an alternative method for this task that takes advantage of the artistic connections between movies. Crew members like to insert small elements (such as scenes, costumes, objects, posters, etc.) in a movie, which reference other movies from the past they have been influenced by. By referencing these movies, they show their appreciation and admiration, and drop hidden messages for attentive audience-members (also known as ‘easter eggs’). The idea behind our new approach is to collect these references, and to build a network where the nodes are movies and the edges are references. Such networks can be studied through algorithms and techniques of Social Network Analysis (SNA) for determining the most important movies in terms of influence.

Our novel technique is based on a combination of several network centrality scores, and determines a ranking of the most important movies in the history of cinema.

In order to determine the importance of a movie in the history of cinema, we usually analyze its success in terms of economic income, that is the amount of money it raised from tickets sold in theaters (the so-called “box office”), or in terms of critical acclaims by means of artistic judgments delivered by professional critics that evaluate a movie from aesthetic and technical points of view.

Recently, researchers in the field of social network analysis have proposed an alternative method for this task that takes advantage of the artistic connections between movies. Crew members like to insert small elements (such as scenes, costumes, objects, posters, etc.) in a movie, which reference other movies from the past they have been influenced by. By referencing these movies, they show their appreciation and admiration, and drop hidden messages for attentive audience-members (also known as ‘easter eggs’). The idea behind our new approach is to collect these references, and to build a network where the nodes are movies and the edges are references. Such networks can be studied through algorithms and techniques of Social Network Analysis (SNA) for determining the most important movies in terms of influence.

In our study we apply a novel technique to the network of references between movies. That technique is based on a combination of several network centrality scores, and determines a ranking of the most important movies in the history of cinema. Our network is built of public data from IMDb (Internet Movie Database), an online database storing information about films and TV shows. The limit of this powerful tool is that it is highly biased towards European and North American productions, so our findings can be considered relevant principally for Western culture.

Then we went a step further and used our movie ranking system for evaluating the careers of key personalities in the movie industry, i.e. directors, actors, and actresses. Our technique ranks personalities according to the number of appearances (for actors and actresses) and film directions (for filmmakers) in top-ranked movies (based on our ranking). This method is based on the assumption that on the one hand the success of a movie is partially dependent on the influence of personalities involved, while on the other hand successful movies boost the career of individuals involved.

We used our movie ranking system for evaluating the careers of key personalities in the movie industry, i.e. directors, actors, and actresses.

In addition, we present statistical analyses of the relationships between film genres, countries of production, and years of release, discovering interesting patterns, especially regarding the collaborations between countries. Moreover, we identify prominent personalities in individual categories (such as a particular country or genre), with surprising results. For example, we noticed that Japanese monster movies (the so-called “Kaiju”, like Godzilla) released in the 50s have a great influence on Western cinema.

Furthermore, top-ranked actresses tend to be less influential that their male counterparts, providing further evidence that the film industry has gender inequality problem. The only exception to this were ‘musical’ movies, where results show a moderate gender equality, and movies produced in Sweden, where actresses are ranked higher than their respective male counterpart.

Our techniques and findings are intended for film historians and researchers in the art domain, but also for cinema lovers who want to re-discover forgotten pearls and stars of the silver screen.