I have been reading about Jess and would like to use it in my project but I
am not sure if it is a right fit or not.  Any help would be greatly
appreciated.

My parameters

I have a large number of users (from a few million to 100million maximum)
I have at most a few hundred rules that generally deal with attributes of
the user
Rules are  are generally about atributes of the individual users but I could
also have rules that are about different but related users
There is a large number of almost continuous per user attribute changes.

Processing can be partitioned to a reasonable number of off the shelf high
performance processors.
Processing most likely can be partitioned in a hierarchical manner such that
we can distinguish between users close to edge cases (where their rules are
almost ready to be fired) and users that are not in the range.


Should I be looking further into Jess for my rules engine? or do I need to
roll out my own specialized software?

Are there any studies that shows the performance of Jess in cases where you
limit the number of rule to a few hundred optimized rule but large number of
data changes?

Thanks,

Daryoush

Reply via email to