stephaniemanuel@fico.com

The better half of decision modeling: Business Terms

Blog Post created by stephaniemanuel@fico.com on Mar 23, 2016

Business terms are the “better half” of decision modeling because without them, you would basically just have a diagram of a decision process, which gives you an overview of the whole thing, but the business terms provide you with the actual information you’re making the decision about. Which, in my opinion, is the fun part.

 

So, what exactly is a business term? A business term represents a piece of information (data) you need in order to make a decision. This information is usually a fact about something you are making the decision about. For example, if you are evaluating a customer, their age or the date their account was opened might be relevant business terms. Business terms are like a list of objects you make your decisions about. Pretty simple, right?

 

Business terms are used in a decision model in a few different ways:

  • As the input in order to make a decision.
  • In the decision logic, which provides explicit instructions for making the decision based on the values of the business terms being evaluated.
  • As the output, which is based on the decision that is made.

 

In a perfect world, everyone (individual or company) would define things in the exact same way. Unfortunately, this isn’t the case and managing all of the business terms for your decision service can get complicated when they come from several different sources. These sources can vary - it may be from one of your coworkers or from someone in a different department or from a different company altogether - and these sources may define a term differently than how you want to define it in your decision service.

fico.png

 

There are two different types of relationships between business terms: equivalent and derived (Note that I named these relationships myself – there is not an official term for these relationships, at least that I know of).

 

The equivalent relationships between business terms are pretty straightforward. They are 1-to-1 mappings of the business terms. For example, your coworker has the business term “carColour” but you’re using “carColor” in your decision service. These business terms are referring to the exact same piece of information, but they are defined (in this case, spelled) in a different way. This is kind of like translating a word from one language to another. 1-to-1. They are equal to each other.

 

The derived relationships between business terms are a little bit more involved. They require some calculation. For example, one of your business terms is “age”, but the data you received from a different source doesn’t have a term for “age”, but they have “dateOfBirth”. In this case, you would need to create a calculation (such as: age = currentDate – dateOfBirth) to relate these terms.

 

To help simplify the organization of business terms, FICO has created the Business Terms Manager! You can upload or create all of the business terms you need in your decision service to the Business Terms Manager and then create relationships between them. The Business Terms Manager is helpful for a few reasons:

  • It is a centralized place to store all of the business terms you use in your decision service.
  • When you have a set of data from an external source that doesn’t match the business terms you have, you can relate them by using calculations.
  • You can keep all of the “messy” relationships between business terms out of your decision service, so the only calculations that are made in your decision service are related to the decisions you are making.

 

If you are interested in learning more about the Business Terms Manager, it will be available soon in the FICO® Decision Management Platform (DMP) 2.0. Check back for more information!

 

I hope I was able to help you understand the basic concept of business terms and their importance in decision modeling. If you found this post helpful/unhelpful, interesting/uninteresting or have any other thoughts, let me know in the comments below.

Outcomes