Much has been covered about including “Rule Engines” in modern BPM packages, but not all rule engines are created equal. Some rule engines are similar in concept to those commonly referred to as “expert systems” and as a result are both more powerful and more complex than rule engines that are simply externalized functions or code. I’ll simplify the concepts and explain why you might want to factor in the types of rule engines a particular BPM package uses when making a purchase decision.
A note on using terminology: In this article, I will use “expert systems” to describe more powerful and more complex rule engines, but use “procedural rules” to describe lighter, procedural rule engines.
The definition captured on the first line here (source: Wikipedia) is a perfect summary of the features that distinguish expert systems from procedural rules engines. Simply put, while expert systems are designed to select a set of information from a range of possible outcomes at once, much like a human, procedural rules engines are told to consider only one possible outcome at a time. The benefit of the former is simplicity, while the benefit of the latter is simplicity, while it promotes thinking in a less structured way, hence a wider universe of possible outcomes.
BPM and Rules Engines: Real-world distribution
Here is a real world example… Let’s say you are a modern quantitative hedge fund and want to evaluate a number of possible securities trading strategies given pricing and baseline data for stocks of a particular industry. These possible strategies can be tens or hundreds (or more). If you are using a more powerful rules engine, you “add one or more facts” to the rules engine and make a set of results or decisions by running all the rules that apply. The rules engine thinks more like a human and takes into account all the possibilities coded into it. If you’re using a less complex, procedural rule engine, you’ll need to code in the execution of each rule (typically configured using an IF… THEN – or similar – structure), which would likely generate more voluminous code. But on the other hand, your logic is simpler and easier to understand for those who are not used to working with expert systems.
As with all other decisions about which system you should choose for your business, the decision depends on your specific business needs. Are you a quantitative hedge fund similar to the theoretical hedge fund described in the previous paragraph and do you have a lot of possible trading strategies to consider for every set of pricing data you receive? If so, you’ll likely want to weigh heavily on whether the candidate BPMS has a full-featured, built-in expert system. On the other hand, if you are an online retail company with a limited number of possible discounts and shipping strategies, using an expert system can be negative due to the added complexity and the learning curve associated with learning to think for your staff. in a necessary way to develop a set of rules for an expert system.
In summary, not all “rule engines” are created equal. Carefully consider the relative strengths and weaknesses of your prospective BPMS vendors ‘rule engines’, whether expert systems or procedural rules engines, and be sure to weigh this information against your specific business needs. Whatever your decision may be, Princeton Blue, Inc. It can help you make the right decision and implement your BPM solution in the most cost effective way; Please visit our website at www.selensoft.com and ProcessMaker BPM page.