An Empirical Investigation on Movie Industry from 1980 to 2018

One of the critical problems in the emerging field of computational social science is how to achieve socially generated “big data” to build a “cleaned” dataset for accessing information about collective human behavior and. The movie is an essential cultural product and watching a movie is one of the most popular entertainment ways in our daily life. A favorite movie can attract the millions audience. Additionally, after years of development, the film industry has produced a large amount of data that can be supplied to the research about collective behavior propagation. Hence, movie industry is suitable candidate for investigating collective behavior propagation in social network. In this work, we construct a dataset including almost all the movies released in the United States from 1980 to 2017. By analyzing the US box office data, we found that “action”, “documentaries” and “drama” are the most favorite movie genres in the past 38 years. We also find that the final total global gross of a movie is stronger related to its best weekly rank, namely, the global gross of the movie is almost proportional to its best weekly rank. Furthermore, we find that although the total box office is rising every year. However, with the birth of the Internet, fewer and fewer people are willing to go to the cinema to watch a movie. Our analysis can help understand how a movie become popular and help movie distributor to public audiences' taste and safety develop a plan for providing movie products.