Context in Industrial data visualization

Get a better understanding of what is context in industrial data visualization

Figure 1. Downtime duration of machine parts.

About context in data visualization

Understanding context in dashboarding is essential and sets you on the path to success when it comes to creating visual content [1].

Today, we will cover this important topic in our journey towards data visualization.

Actually, data is visualized since a long time if we don’t consider technologies used for it nowadays. In fact, Physical artefacts such as Mesopotamian clay tokens (5500 BC !), Inca quipus (2600 BC !) and Marshall Islands stick charts (n.d.) can also be considered as visualizing quantitative information [2].

Figure 2. Mesopotamian clay tokens. Denise Schmandt-Besserat

Even then, data was visualized to point out information through its content and context.

Content in data visualization is the measure. For example, we have a battery producing energy. Energy is our content. Content itself is not relevant in Data visualization. A dashboard indicating only the energy a battery produces does not contribute into decision taking. That is where, context comes into play.

Context ( from Latin contextus, con-’together’ + texere ‘to weave’.) means “the construction of”. So context is the environment containing our content.

Lets see together how to use context when creating our dashboards.

Context

Understanding context should not be underestimated. Indeed, it is the starting point when creating dashboards. We begin our visualization process once we are sure we got clear context. Our process starts by identifying the context in our data source’s columns.

In industrial data visualization, we use explanatory analysis. Our Dashboards should explain clearly our data to our audience.

Knowing our audience improves the quality of the message we are willing to share throughout our Dashboard. The more specific our target audience is, the more our Dashboard will be significant to them. For instance, in Figure 3. below, our audience is surely not our marketing department.

At this point, we are the master of our data. After presenting a Dashboard, our audience must be able to take a decision based on our data. We need to ask ourselves questions such as “ who, what, when, how” in order to decide our context.

Of course, we need to check if our context is available in our data source and is meaningful within our content.

A few examples

Figure 3. Downtime of machine parts.

In our example above, I see downtime duration of machine parts. Our context here is “machine parts”. Let’s assume that our audience is a technical support team manager.

So, I can explain to him that the packer had a higher downtime compared to a filter. Eventually, my message here is to point at him that my packer has a high downtime value compared to other parts.

Figure 4. Product prices of product X and Y during a given datetime.

In Figure 4. our context is time. We can notice that we can’t give the same type of explanation by reading this example compared to the one in Figure 3. In this one, I can explain when my product X had higher prices than my product Y.

Indeed, in Figure 3, we have no explanation related to datetime.

Our context will be displayed in our widget. It is also important to adapt our diagrams according to our content and context.

To see an example using content and context, we can check out this article about Sankey diagrams.

Conclusion

We can conclude that context in industrial data visualization is crucial. It is the As of Heart that determines the clearness of the message we wish to communicate.

References

[1]Cole Nussbaumer Knaflic. Storytelling with Data : A Data Visualization Guide for Business Professionals. Hoboken, New Jersey, Wiley, 2015.

[2] Schmandt-Besserat, Denise. How Writing Came About. Austin, Tex. Univ. Pr, 2006.

Figure 2. Schmandt-Besserat, Denise. How Writing Came About. Austin, Tex. Univ. Pr, 2006.

Knowledge enthusiast. Mostly about science, education and rhetoric. Writing about data viz in dashboards.