Professor Ramesh Sharda of the Oklahoma State University says his software is based on looking at the content and seeing it in a context of movies that have been running in the past.
The software has so far analysed more than 800 films and the results will be published in a paper which appears in "Expert Systems With Applications" journal, next year.
According to Sharda there are seven criteria to tell if a film is going to be successful. These include its rating by censors, competition at the time of release, strength of the cast, genre, special effects, whether it is a sequel and the number of theatres it opens in.
The software then uses a neural network to process the results. Films that make less than a million are a flop while a "blockbuster" earns more than $200 million.
Sharda's software is apparently dead-on 37 percent of the time and pretty close 75 per cent of the time. I would have thought that if Hollywood ever uses the software, it will mean that films will stay the same because new types of films will not be tried.
More here. ยต