Extending Bayesian Filters to Personalization

You log into your email account. You have multiple folders - you realize that the emails from some Mr. X (the hot guy you met at a friend's birthday party) have been transferred to the 'Personal' folder.

You get up one morning. And see that your microwave oven has already 'switched itself on' now that it knows that you wake up every morning at 6 o' clock.

You insert your ATM card and realize that the appropriate option is selected according to your preferred patterns of operation.

Probabilistic reasoning works well in a large number of areas - the probability of Karan Johar movies being loud and emotional, the probability of me spilling my coffee, the probability of Meer ( a classmate and a good friend.) getting late for college and the probability of Janardan laughing loudly, they are all very high.

So, could we design a personalized coffee holder that helps me hold my cup tightly? Could we design a personalized alarm clock that keeps irritating Meer so that he finally learns to reach college on time?

That's how Bayesian filters can be extended to personalization.

Name of the Study Paper: Extending Bayesian filters to Personalization
By: Aaditya Joshi, Mihir Kachalia, Janardan Kelkar
Under the guidance of: Prof. Mrs. Kavita M Kelkar, Dept. of Computer Engg, KJ Somaiya College of Engineering, Vidyavihar.
Status: Work in progress.