What can Venn diagrams teach us about doing data science better?

Abstract Data science is about deriving insight, learning and understanding from data. This process may be automated via the use of advanced algorithms or scaffolded cognitively via the use of graphs. While much emphasis is currently placed on machine learning, there is still much to learn about the...
Ausführliche Beschreibung

Abstract Data science is about deriving insight, learning and understanding from data. This process may be automated via the use of advanced algorithms or scaffolded cognitively via the use of graphs. While much emphasis is currently placed on machine learning, there is still much to learn about the role of the data scientist, in particular the thinking process by which he reaches conclusions. The thinking process of the data scientist needs to be scaffolded as the human brain is easily overwhelmed by many variables. Graphs are a form of data abstraction and constitute an essential part of the data scientist’s toolkit. Graphs are also a viable scaffold on which the data scientist may gain familiarity with data. But the process of extracting insight from graphs is not always a trivial or straightforward process; it requires interpretative logic as well. Generalizing from the example of a simple graph type, the Venn diagram, we discuss various logical fallacies that can be committed when interpreting a Venn diagram. Amidst various considerations that dictate how a graph should be tackled, we explain why context is most important, and should form the first guiding principle during data analysis. Ausführliche Beschreibung