Need some problem-oriented examples

I am open to suggestions on new examples to be added to Drools. Do you have any particular problems in mind you would like to see a Drools implementation of? If so, add them as comments to this issue, or better yet, submit the code to an example of your own.

The thing is, I'm one of the Ignorant Masses (tm) that need convincing and educating. I'm not all that sure which problems would make for good examples.

I have been researching the origins of the RETE algorithm that underlies Drools and other popular rule engines, and through that process I have found links to a number of example problems that have been solved with rule engines in the past. It's important to note here that a lot of the earliest uses of rule engines was related to Expert Systems. Here are some places worth checking out:

Expert Systems Benchmarks - http://www-cgi.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/expert/bench/bench/0.html

Contains OPS5 examples related to: Aeronautical Route Planner, Miss Manners, Waltz(DB) [2-D/3-D visualization algorithm], and Weaver [VLSI
routing].

CLIPS - http://www.ghg.net/clips/CLIPS.html

Contains CLIPS examples related to: National Cholesterol Education Program guidelines, Animal Identification Expert System, Automotive Expert System, Cannibals and Missionaries Problem, Farmer's Dilemma Problem, Electronic Circuit Input/Output Simplification Expert System, Monkeys and Bananas AI Planning Problem, and more.

Jess - http://herzberg.ca.sandia.gov/jess

Contains Jess examples related to: Self Learning Decision Tree Program, Indirect Reasoning Problem (Water/Zebras), Damped Algorithm Control Problem (Pumps/Tanks), and more.

Converting any of these OPS5/CLIPS/Jess rule examples into Drools should not pose a challenge if anyone thinks one or more of them would be a good addition to Drools.

I have been racking my brain to come up with a good example to add to Drools to demonstrate something resembling a real-life application of business rules using Drools. One idea I have is to create an example related to the domain of Telephone Call Billing.

It is the norm for phone companies to offer hundreds of different telephone calling plans, each differing by many factors. Factors such as these can affect the price of a single phone call: time of day (peak hours), day of week, day of year (holidays), local calls, interstate calls, international calls, in-network calls, roaming calls, calling circles, minimum call time, minimum usage charges, taxes, etc. Furthermore, depending of when a customer signed up for a calling plan (for example, during a promotion), rates can differ from one group of customers to the next for each aspect of the calling plans. And let's not forget about scenarios where a call crosses a peak-hour time boundary, or a cell user crosses a roaming boundary while in the middle of a phone call.

In support of billing calls, phone networks can produce Call Detail Records (CDRs) which record details about individual phone calls – details such as call origination, call destination, date/time of call, disposition (aka answered/no_answer/busy), etc.

As a consequence of all this, the process of determining the proper amount to bill a customer for each call can be a daunting task.

It seems to me the problem of determining the rating/billing of individual CDRs is a good application of a rule engine.

Any nice example for using drools in real life? Like the Andy mentioned. It's been so long since the last comment and issue is not resolved