Futurist: Diving into the Data quality pool

The overarching question when considering the concepts required to build a good knowledge management system is the reality of data quality. Data quality can mean many different things. In the KM world, it can be the difference between success and failure. We have to be careful that the information shared by the KM system is of acceptable quality. No, that’s not grammatically perfect. It is the text that solves the specific problem that the user is inquiring about.

?

By design, KM systems expedite information retrieval and utilization, meaning the data retrieved must be of the highest quality. Again, this is not grammar. These are the actual core ideas within the knowledge that is acquired. However, there is a huge question regarding data quality. The question potentially drives data quality more than the reality of the data itself.

?

What is the distribution of resources in building a human design, machine, intelligence, or HDMI system versus the amount of effort required to optimize the quality of the data consumed and presented to end users? That question becomes very painful, but there is a formula that you can use. Data quality presented in a KM system is legitimately focused on the concept of the cost of the problem minus the cost of retrieval, resulting in the net financial benefit of optimized data. Or, more bluntly, what is the quality of the data you need to return in a specific ask?

?

It is important to note and consider several parameters. One of them is human life. Data quality is more critical if a human‘s life is at stake. Wrong information or bad information, when applied to the life or death of a human, is a bad thing. The same should be true for all animals living on planet Earth. So, data quality is critical if it’s a life-or-death decision. That doesn’t mean organizations responding to an RFP don’t need to worry about data quality. The reality of producing a response to an RFP is that the data needs to be of good quality for the organization.

?

That is what human-designed machine intelligence (HDMI) elements you consider in your KM system. Focusing on data quality becomes more important if you intend for the KM system to provide some data quality components. However, the actual HDMI system has to have high-quality data as its initial input, or you risk receiving garbage. The old garbage in is almost always garbage out, so as we build an HDMI-driven KM system, we have to consider where and what we will do for data quality. Many years ago, I released a book called Transitional Services. In that book, I created a human-driven knowledge management system. I called it digital lifecycle management or DLM. In the DLM system, direct human SMEs retrieved, evaluated, and promoted known good solutions to problems. Today, in reviewing that system, I would instead have HDMI agents monitoring the system for heavily used intellectual capital or IC. Use is an important component. Use points to the human validation that the information has value. The HDMI agents would surface the most used responses to a specific question or problem in that scenario.m

?

We would then have a mixture of experts to evaluate the presented data. We would have some experts. Let me lay out a few postnatal experts for the system. A list that is not comprehensive is simply a starting point. Experts connected to the data system. Each expert would operate independently. The Experts might find the original solution, propose alternate solutions, and evaluate solution quality. A mixture of expert structures would allow for creating a data quality engine. The mixture of experts verifies and validates that system Information, providing it back to the user within additional days.

?

Of course, you still have to answer the question of how much money you intend to spend on data quality in the short run. We organizationally have to be evaluated against the cost of bad information getting out. Unfortunately, that is a more esoteric argument in that a chunk of the impact of bad information leaving your company is the reality of reputational damage. Measuring the financial impact of the concept of reputational loss is very difficult. Every organization needs to have a data quality initiative. The question and argument then have to be whether the data quality impacts our organization, so we need to change our pursuit of data quality. So, as you and Barb are building a KM system focused on delivering knowledge management structures around your intellectual capital, you have to ask yourself what is the risk of people delivering bad data outside our company.

?