what are your features? what is your training dataset?

naive bayes assumes that the attributes are independent of each other (i.e. the joint distribution can be written as the product of individual distributions)

exactly, the assumption is correct. attributes are independant of each other.

features -> not much , i'm planning on taking probabilities of each word in the sms being a spam word or a ham word , based on the existing dataset.(dublin University)

Now that i got the hold of it, i'm seeking for some modern techniques apart from Naive Bayesian, since it is the most basic one in filtering n detection.

hi

each parameters are the attributes in naive Bayes.

naive Bayes will calculate the conditional probability for each of them.

Then it finds the probability of occurrence of one of the class for a given instance  i.e., your new mail.

That's how it works.

Hope it will help you.

Best

Dr. Indrajit Mandal