Though content has a foundational role in knowledge management, content remains a largely hidden actor. It can be quite disconcerting to review the program of conferences dedicated to, among other things, knowledge management, and not hear the term “content” spoken in any of the presentations. It’s as if content is the love child that hovers around the edges of the family, a definite presence but never acknowledged.
Information vs. data vs. content: understanding the key difference
Whether one chooses a dictionary definition of knowledge management as “efficient handling of information and resources within a commercial organization”, or an academic definition of “the process of capturing, distributing, and effectively using knowledge”, a common omission is the recognition of content as a primary actor.
The term most often used to refer to content is data, though data is definitely not the same thing. Data can float, unencumbered by context, inserting itself into any conversation to add specificity, depth, and texture. “The markets are down” has less context than by adding a data point: “The markets are down by 75%.”
Starting with data: the foundation of enterprise knowledge
Data comprises the building blocks of content – that is, a single data point can be an accurate representation of a concept, but contains no context that helps humans derive meaning. For example, 353 323 428 800 could be a part number, a bank account number, a telephone number, a lottery ticket number, or similar. Without sufficient context, human comprehension is limited.
Turning data into meaningful content
Add some context, and you have content. Content can be defined at its core as human-usable, contextualised data. That means that we add enough context to the data for humans to decipher its meaning. Going back to our mystery number, we could declare what the number represents. “This number is an IBAN” would be a basic example of content. Some people may not know what an IBAN is for, but now they know not to use the number as a telephone number or a part number.
Content is the primary medium that humans use to consume information. That’s the case whether it’s visual content, such as images or videos, or text-based content—everything from social media posts to academic papers, from business reports to product instructions.
Context-enriched content: the key to actionable information
The transition from content to information is the addition of more context. As previously mentioned, understanding a number is an IBAN may need some additional context to turn that piece of content into information. “This is an IBAN, which is a transit number used by banks when transferring funds between countries.” Through more context, content is turned into actionable information.
The content layer: a crucial step
Ways of describing knowledge management have long skipped over content as a layer of understanding, and gone directly to information. That’s true whether the reference is the traditional DIKW pyramid or the KM Continuum.
Going directly from data to information by skipping over content wasn’t previously as crucial an omission as it is today. There are a couple of intertwined reasons that a separate and distinct content layer is important.
One crucial reason to distinguish content from information is that naming concepts allows us to broaden our understanding of the use of the named things. This acknowledges that content has unique needs from data, such as positional encoding, grammatical structure, cultural nuance, and so on. This is distinct from the role of data as an indicator of specificity, and the role of information as potential insights.
Content is nestled squarely in the middle, taking on the role of capturing facts that can be combined and supplemented to form information.
The rise of microfacts: surfacing business intelligence in the AI era
Now, one knowledge management trend is to enable service portals for product support. In a nutshell, the idea is to consolidate all product content into a single repository that, when queried, serves up an accurate answer.
In reality, it’s a bit more complicated than that – the single repository may actually be several repositories and databases, and the answers may not be the information someone is looking for. The end users may be customers or prospects, or they may be service agents or repair technicians, but the goal is the same: find and serve up the exact piece(s) of content queried through the use of a chatbot.
What is a microfact?
As a result, we are seeing the rise of content in the form of microfacts. Microfacts are short pieces of content that contain specific, targeted facts. They are suited to chatbots because they are short enough, specific enough, and contextual enough to be retrieved upon query. They have more context than data, and when combined with other microfacts, provide sufficient context to the user to be actionable.
Going back to the IBAN example, a microfact could be “in Ireland, an IBAN contains 22 characters” or “an IBAN begins with a two-letter country code”. This piece of content is an isolated fact, but combined with other microfacts, becomes information.
These facts can be served up sequentially or through an LLM, combined into a single, actionable statement, such as “In Ireland, an IBAN contains 22 characters, beginning with a 2-letter country code, followed by 2 check digits, a 4-letter bank code, a 6-digit bank identifier, and ends with your 8-digit account number”.
Getting microfacts right: why the content layer is critical
Microfacts help ensure that a chatbot can retrieve, process, and present information accurately and reliably. However, a lot of this depends on the quality of the microformats and how they are developed. Some organisations may create microfacts as independent pieces of content; the appeal of this approach is the seemingly easy creation and maintenance of the content. Other organisations use metadata to isolate microfacts while keeping the microfacts in place within their modular topics.
As the adoption of AI rises, so too does the need for microfacts to surface the right information at the right time, to turn it into explicit knowledge. This, in turn, helps to surface business intelligence that, as content is combined and recombined with each query, can reveal business insights, facilitate knowledge retention, and support decision-making. It’s important, then, to lay a solid foundation, by including content on the stage with the other actors in the knowledge management framework.