In the tech industry, we also have our well-known “Coke vs. Pepsi”, “Avis vs. Hertz”, or “Mac vs. PC” debates. In the automated decision management category, the question that keeps coming up is “data vs. knowledge.” The aim of this blog post is to show that from a practical point of view, data and knowledge can be found in the same application. To do this, we will show it with SMARTS, Sparkling Logic’s decision management platform that allows users to combine data and knowledge without them entering the “data vs. knowledge” debate.
Origin of the debate
The “data vs. knowledge” debate dates to an old debate about whether knowledge about a subject should be hand coded or machine learned. A first camp of researchers and practitioners sought to encode this knowledge in the form of rules and an inference engine that runs on these rules to supply answers to user questions. A second camp sought to develop programs that learn from available data using statistical methods to generate models that can make predictions from unseen data. At Sparkling Logic, we support a pragmatic approach that consists in using data, knowledge, or both depending on the problem and the situation at hand.
It is all about the situation
There is no such thing as a stand-alone decision management application. It is often built with the purpose of being integrated into a larger system for loan origination, risk management, product configuration, or other similar applications. As I wrote before, there is no one single approach. It is all about the situation.
Data is everywhere, easy to collect, organize, and transform into predictive knowledge. So, if you have a lot of data, it may be better to build your decision management application around that data if the new observed data does not deviate too much from the old, learned data.
On the other hand, when you have knowledge whether in the form of rules or procedures, it is better to build your application around this valuable knowledge if it is easy to capture and code into the application.
If you have both data and knowledge, why not using the two, when you can do so in a modern decision management platform such as SMARTS, the subject of the next section.
The SMARTS way
SMARTS is a decision management platform that enables creating, testing, deploying, and improving automated decisions in an integrated platform. I will not detail it here, but you can find a brief overview of SMARTS on our blog page and a full description on our resources page. Instead, I will focus the rest of this article on how to use SMARTS when you have plenty of data or domain knowledge about the application you want to develop.
You have plenty of data
For situations where you have plenty of data, SMARTS proposes two tools: RedPen and BluePen.
With RedPen, you write decisions in the form of rules using a use-case driven approach. A loaded data sample supplies the context for the rules and enables immediate execution and testing of each rule. RedPen mimics what subject-matter experts do when they flag decisions.
When you activate RedPen, you can pin an existing rule, a field of this rule, or a rule set and change it as if you were using a real pen on real paper. You can also create new rules with RedPen, SMARTS automatically turns them into executable rules.
On the other hand, BluePen lets you explore and analyze your data using your domain knowledge to find predictors. Then, using these predictors, you can generate a model in the form of legible rules and integrate them into your decision logic.
Using BluePen, you can engineer or change the models any time you need to. Without heavy investment in data analytics tools and efforts, you can evaluate BluePen models in simulations and quickly deploy them in the context of an operational decision.
You have domain knowledge
For situations where you have knowledge, SMARTS proposes two additional tools: SparkL and Pencil.
SparkL is Sparkling Logic’s language for writing rules in a natural language fashion. SparkL comes with everything you need to write rules —mathematical expressions, string manipulations, regular expressions, patterns, dates, logical manipulations, constraints, and much more. You can express any imaginable decision logic and symbolic computation, making it the choice for highly sophisticated decisioning applications where the conditions as well as the actions can take a wide variety of forms.
Pencil is our DMN compliant graphical decision design tool for uncovering, documenting, and sharing decisions with colleagues and partners. With Pencil, you just drag and drop graphical shapes to form a complete decision diagram. Then you add business logic to the graphical shapes and let SMARTS execute it.
Pencil helps you think about the ultimate decisions in a structured way, starting from the top-level decision to smaller sub-decisions. This iterative process is very friendly and amazingly easy to share with colleagues or partners working on the same project.
In addition to the above tools and language, SMARTS comes with a built-in dashboard to measure and improve business outcomes of the taken decisions.
You have both
If you are lucky enough to have both data and knowledge, you can leverage your models’ outputs by using it as the input to rules. For example, your loan management application could run a model that calculates a score and another model that calculates a risk and use that score and risk in a rule to calculate a price.
You can also do it the other way around, using the outputs of rules as inputs to your models that you would have trained with data. For example, your application might run a rule to classify a loan applicant, then run a model to calculate their risk of default and another model to calculate the price.
Whether you have data, knowledge, or both, SMARTS uses them as sources of information for the automation of your operational decisions.
- Data and knowledge do not have to be antagonistic. They can both be used as inputs to automate decisions.
- SMARTS is a modern decision management platform that enables their combination in an elegant and seamless way. For SMARTS, data and knowledge can be used as sources of information.
- When you have a lot of data, you can use RedPen to write rules without learning a special rule language or syntax, just starting with the data. You can also use BluePen to learn from data and turn it into rules.
- When you have knowledge, you can use SparkL to encode it into rules, from the simplest to the most complex rules that your application may require. You can also use Pencil when designing, documenting, and sharing your decisions are part of the requirements.
- Our mission is to enable customers to implement the most demanding decisioning requirements and to easily change and improve them over time. Whether you have data, knowledge, or both, we can help. Just contact us or request a free trial.