Standing on the platform waiting for the #7 to approach, I inadvertently began a conversation with a craggy gentleman about the controversial meaning of life. A gruff sigh got my attention as the man looked down at his wrist. The train was about 5 minutes late. “What am I doing here?” his lips muttered as he looked up from his watch. I thought for a moment. Does he mean here as in on the platform waiting for the train? Does he mean here in that very spot where he stood sandwiched between the column and the subway performer? Does he mean here in this city? It was difficult to infer the true meaning of his question, so feeling sociable and inquisitive I initiated a conversation. I glanced over with a smile and implied “Rough day?” In hindsight I wish I kept my mouth shut and simply moved over.
This situation prompted me to think about the importance of semantics. Semantics is the study of meaning and interpretation within linguistic expression. How many meanings could the word here have? In this situation at the train station, here required knowledge of additional context. The gentleman provided semantic clues during our conversation that helped me accurately discern between relevant and irrelevant content to understand his meaning.
The conversation moved from the platform to the train -- standing room only, of course. He ranted about politics, religion, healthcare, education -- any topic that perturbed him. They all seemed to meld together into one big ball of anger and discontent. The pace dragged on and my eyes started to glaze. I was lucky if I got a word in at all. There was a pause in the conversation. “Son, do you know why you are here?” he asked. Being a wise guy and looking for anyway to get out of this intolerable conversation, I cocked my head to one side and pointed to the floor beneath my feet and replied, “You mean here, or do you mean here on the train or do you mean here talking to you?” He said “Let’s not argue semantics, you know what I mean?”
"No sir, I’m afraid I don’t."
Semantics Are Critically Important for Business Discovery
So, what does this have to do with Business Discovery? Semantics are the key to turning a deluge of raw data into a relevant source of actionable information. When applied correctly, semantics cannot only define relevant meanings of data, but it can be used to apply a consistent and reusable approach to managing BI applications and its ever expanding sources of data.
When we discuss linguistic semantics we interpret the further meaning of words and phrases by using existing knowledge and our surroundings. Sights, sounds, and symbols can all have a semantic meaning. However, without semantic clues it is left up to an individual’s own interpretation and knowledge, which can be flawed.
The Semantic Layer
BI software providers use the descriptive form of the word “semantic” to describe components of their software. This comes in the form of what is usually known as a semantic layer or framework -- an architecture that describes and provides a consistent meaning of reusable business application attributes, objects, and data. “A layer of meaning,” as I like to call it. This layer tends to be strictly defined by BI administrators who should have familiarity with the business problems and strong collaboration with peers, rather than using loose interpretations.
In BI, a semantic layer bridges the gap between the way users work with data and how BI applications process it. This is achieved by the use of metadata. Metadata is an ambiguous term that can be defined in this content as data about data. (how ironic) It is traditionally designed in a top-down fashion. Created in a prescribed manner, it provides a consistent foundation for reuse in development, analysis, and even data governance. Properly defined metadata can reduce information inaccuracies and misguided decisions by decreasing the number of disparately defined objects, metrics, and dimensions used throughout an organization. "Business users get data, IT gets governance." Join me in my next article, “Metadata the QlikView Way,” where I expand on these topics and discuss the value and benefits of defining metadata the QlikView way.