Anti-Spam Tech Reviews Movies

Anti-spam software producer Unspam deployed its email evaluating technology towards an unlikely goal. Instead of working as anti-spam software for Outlook, it predicts if a new movie will fly or not.

Like many other recent email content filters, Unspam uses Bayesian analysis. Bayesian analysis allows for easy automated evaluation of large data sets. Instead of feeding their engine information on spam, they fed it information on movies, and whether or not they were successful.

The fuzzy logic technology behind SpamButcher is similar to Bayesian analysis, but is more likely to use a dataset hand-tuned by a human. Both approaches have shown good success in spam fighting, and have their own pros and cons. Adding large amounts of new data to a fuzzy logic filter can be more time consuming, however the process is less likely to result in, "red-hearings" which can lead to false-positives.

Variables regarding both the film itself, and the time it was shown were considered. It is possible that "favorites" in prior years were likely to secure certain desirable times beforehand.

32 different films where processed based on hundreds of data points. In the end, the electronic movie evaluation system's judgments mirrored the actual winners of the Sundance festival. Considering the relatively large number of films being considered, the selections are likely not a coincidence.

Perhaps like the space-race, technologies developed for the purpose of eliminating spam can benefit other disciplines. Using the latest advancements in A.I. to block email in Outlook may not be as lucrative as trying to predict the outcome of an election or horse race.

Back