Bayesian filters are created from scratch each time a new message comes. A
caching mechanism is needed like SvmManager class. Only one instance of a
filter should be available and this filter should be trained before obtained.
Filters are needed to be populated after or before scheduled svm training.
Original issue reported on code.google.com by aykutuy...@gmail.com on 11 Jun 2011 at 7:16
Original issue reported on code.google.com by
aykutuy...@gmail.com
on 11 Jun 2011 at 7:16