As I write this, everyone is freaking out over COVID-19. We know it is a highly contagious and dangerous disease, with particular risk to groups in a confined space such as an airplane or cruise ship. Researchers are scrambling to model the contagion risk, but detailed data is almost nonexistent. With a good model, you could create a decision service that would output the likelihood of a passenger contracting the disease if on a flight with an infected patient.
In recent posts I reviewed the use and benefits of BKMs and contexts for DMN modelers and stakeholders. This time we'll continue that theme with an explanation of another of DMN's woefully underutilized features, decision services. The introduction of decision service as a formal element in DMN came about in a surprising way. The spec's only real example of a decision model, a lending decision, featured a DRD meant to be executed in two or three steps.
By:
Bruce Silver
February 24, 2020
dmn
Read More
Last time I showed the value of Business Knowledge Models (BKMs), those misunderstood and often-maligned elements in DMN diagrams. In this post I'm going to try to do the same for contexts, another DMN feature that is similarly underappreciated and frequently disparaged by tool vendors that don't support them. Unlike BKMs, contexts are not DRG elements, meaning distinct shapes in the DRD. Instead, a context is a type of boxed expression, a standard tabular format for the decision logic of a decision or BKM.
In a recent post, I showed how DMN could be used to model and execute decision logic on complex XML input data such as loan application information in MISMO format. The MISMO schema was envisioned as an enterprise data dictionary for the mortgage industry, standardizing the names of various information elements for use across any mortgage-related application. As such, however, it is effectively impossible to use MISMO data directly in decision models.
Newcomers to DMN may wonder, What is the point of Business Knowledge Models (BKMs), those funny-looking boxes with clipped corners in the DRD? When I was first introduced to DMN, struggling to understand the DMN 1.0 spec, I wondered the same thing. The spec was (and remains) largely impenetrable to decision modelers, except for the example model in Chapter 11. That example has a BKM attached to every decision in the DRD, but for no obvious reason.
Fannie Mae is one of two major Government Sponsored Entities (GSEs) underpinning the residential mortgage market in the USA. They supply liquidity to the system by purchasing mortgages from lenders and repackaging them as interest-paying securities sold to investors. Without them, we wouldn't have the low-rate long-term fixed rate loans that so many home buyers depend on. Recently Fannie published a white paper titled "Without Data Standards, the Mortgage Industry Doesn't Go Digital.
When BPMN 2.0 was developed a decade ago, the task force’s primary goal was making BPMN models directly executable on an automation engine, something that wasn’t exactly possible with BPMN 1.x. Consequently, the rules of the BPMN spec focused almost exclusively on “operational semantics.” In doing so, they lost sight of what the majority of BPMN users cared about much more, which is making the process logic clearly understandable from the diagrams.
While DMN adoption continues to accelerate, we can only admire the current frenzied interest in machine learning. Both technologies are trying to do similar things - make decisions - but they go about it in very different ways. DMN, which evolved from business rules and, before that, expert systems, is based on intuitive understanding of the logic. Decision modelers are assumed to be subject matter experts. On the other hand, machine learning - formerly called predictive analytics or data mining and now marketed as "
DMN is excellent at working with tables of data. Such a table in DMN is called a relation. A relation is defined as a list of rows with a specific number of named columns. The number of rows in a DMN relation is unspecified. Most decision models work with tables in this form. But what if you want your DMN model to work with numeric tables having an unspecified number of columns and rows?
Practical DMN: The Basics Want to learn DMN decision modeling without training? DMN Method and Style, 2nd edition is a good place to start. But to really learn it, you need to go beyond books. You need to get your hands dirty in a tool. I suggest starting with a free trial of Trisotech DMN Modeler. When you log in, go to the EU-Rent repository (File/Open/EU-Rent) and open the model called EU-Rent Pricing.
By:
Bruce Silver
September 11, 2019
dmn
Read More