Information and data—two words we use almost interchangeably and without agreed definition in the IT industry. Add architecture, management or governance to either or both of them and the resulting cognitive dissonance can reach fever pitch. Speak to anyone in the business about some or all of the resulting six terms and you can be sure to find confusion and overlaps ranging from the unimportant to the seriously debilitating. So, you may ask in all fairness, what I mean by information architecture in a general sense and as applied here as a practice name within the Eckerson Group.

Let’s start with the confusion between data and information. I devoted more than half a chapter in my recent book, “Business unIntelligence—Insight and Innovation Beyond Analytics and Big Data”, to the topic. While one can approach the distinction from many directions—history, philosophy, etymology and more—I chose to be pragmatic. Information, therefore, in its broadest sense is what we use to communicate with one another. It has a physical existence in the form of spoken or written words and sentences, sounds and music, as well as pictures, both still and moving. It may be fact or fiction, past tense or future perfect, commands to do or cries for help. Information is the recorded and stored symbols and signs we use to describe the world and our thoughts about it, and to communicate with each other. Today, information is increasingly digital, but also includes paper, books and analogue recordings.

Data, from its root as facts or measures, is therefore a subset and derivative of information. Since the proliferation of computers in the 1960s, the word data has become synonymous with the “stuff” that resides on computer disks and is processed by applications. Data is thus a simplified and well-structured subset of real-world information that is particularly suitable for processing by digital computers. Data modeling is one of the primary ways in which data is derived from information. And, even here, we see the confusion: we actually model the information needs of the business to derive data structures, so information modeling would probably be the more appropriate phrase. Modeling deconstructs information into two parts. It creates a meaningful, persistent structure in which values or instances can be stored. The former is often called metadata and the latter data. Metadata, in the context of what we’ve just discussed, is a misnomer; more correctly, it is information about the context in which the actual data can be used. Today, I prefer the term context setting information.

And, having settled on these distinctions, we can begin to construct a replacement for the old Data-Information-Knowledge-Wisdom (DIKW) framework that dates back to 1989 from Russell Ackoff. This modern meaning model (m³) is shown in the accompanying illustration, where information, including data, comprises the physical locus—what is actually instantiated in some physical medium.

Now, setting aside management and governance, let’s finally return to information architecture. By architecture I mean, of course, conceptual structure and logical organization. Here, information spans the entire range of material that can be stored digitally, from traditional operational and informational data, through metadata, all the way to what is often called “unstructured data” or content, including big data and data generated in the Internet of Things. When considering information from this viewpoint, its ultimate human source and destination is paramount. These are shown in m³ as knowledge and meaning, the mental and interpersonal loci respectively. The remit of the information architect, therefore, is to take the human view of the world and map it to something that can be stored and processed in computers and vice versa.

With the ever increasing breadth and depth of information that is becoming available digitally, and the vast strides being made in analytics and deep learning, a comprehensive approach to information architecture is becoming central and mandatory to the design and delivery of new information systems that are fit for purpose, performant, ethical and humanized. I shall delve deeper into these aspects in part 2.