-->

KMWorld 2024 Is Nov. 18-21 in Washington, DC. Register now for $100 off!

Trees, chains, and brains

Article Featured Image

The inherent flaw in this approach is that the outcomes are expressed as expected, i.e., average, values, often based on subjective probabilities. The good news is that with recent advances in data analytics, much of the subjectivity is being supplanted with “data-driven” objectivity. This includes using decision trees as the underlying structure for many machine learning models.

Role for KM: Minimizing garbage in equals garbage out. By combining data analytics with human expertise and sensemaking, KM can help improve both relevance and accuracy in predicting outcomes. However, this type of analysis is mostly based on correlation, which can lead to poor, and in the case of novel situations or “black swan” events, very poor, decisions. Which brings us to ...

Chains

As in event chains, including Markov chains. We’ve seen how decision trees estimate the chances for the occurrence of possible outcomes of one or more decisions (go/no-go) as impacted by outside events (rain/sunshine). Event chains, in contrast, show the cascading effects of a series of events from start to finish (how one leads to another). A common example is a latent error in the requirements definition phase of a project that ultimately results in significant cost and schedule overruns downstream.

Simulation plays an important part. Probability estimates, such as that 30% chance of a washout in our decision tree example, are typically generated using a chain of cause/effect models/simulations. This is how one method can feed another.

Role for KM: As complexity increases along with the number of variables, such as in the broad fields of climate, pandemics, monetary policy, geopolitics/ war, etc., somebody needs to track and manage the myriad data and knowledge sources involved. Knowledge graphs, anyone?

Brains

For brains, read “intelligence.” More specifically, “decision intelligence.” The term was first coined in the early 1990s. However, a structured methodology didn’t appear until around 2010, when it was introduced by Lorien Pratt and Mark Zangari. The methodology consists of five steps:

1) Decision requirements, including carefully framing the decision;

2) Decision design and modeling;

3) Decision reasoning (applying simulation and assessment tools and methods, including risk assessment);

4) Decision action (clear communication, execution, and monitoring); and

5) Review (good old-fashioned lessons-learned).

Best of all, it ties everything into a single structure, called a causal decision diagram, or CDD. You can learn more in Pratt’s and Nadine Malcolm’s Decision Intelligence Handbook (O’Reilly, 2023).

Role for KM: Some assembly (and lots of facilitation) required, which means innumerable opportunities, especially when it comes to building and reifying those CDDs. This includes managing the flow of money and resources; identifying and mapping a potentially vast ecosystem of people, organizations, and systems, along with their interrelationships and interdependencies (including all the technology platforms supporting the decision processes); external (i.e., uncontrollable) influences; and outcomes (measuring how effectively and efficiently the requirements were met). And so much more, including those lessons-learned mentioned in Step 5.

To gain a greater appreciation of what’s involved and how KM can help, check out the sample CDD developed during the COVID pandemic, along with other interesting examples (github.com/quantellia/di-handbook-supplemental-materials/blob/main/dihb_0002.png).

Do try this at home

According to Quantellia (quantellia.com/Data/HighPerformanceDecision-Making.pdf), 86% of organizations don’t follow a formal decision-making methodology. Don’t let yours be one of them. Start today by thinking about an important decision you or your organization is facing. Build a CDD for that decision, paying particular attention as to where KM could help produce a better outcome. Measure and refine. Rinse and repeat.

Like trains, planes, and automobiles, these trees, chains, and brains are part of our journey to a better future. As the world continues to grow more complex and the consequences of poor decisions become more detrimental, we’re going to need these tools—and more.

KMWorld Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues