Making Data Useful

Now, you have learned why good data quality is crucial for data-driven processes, and that things can quickly go wrong if the data you handle is not complete and correct. If you feed a system with inconsistent, incomplete or incorrect data, you will also end up with results of correspondingly low quality. “Garbage in, garbage out,” as we now know it’s called.

You’ve learned to be aware of biases in the data and what governance really means—and why good management and trust are necessary in data-driven systems and processes.

We have also looked at how to proceed when identifying data and choosing tools. And we have created a checklist for what should be included in a data cleanup job. From that point, you can develop the data from insight to value: A bit like how a data scientist operates!

Another thing a data scientist deals with is asking good and relevant questions. Gathering information and thorough preparation are indeed things that occupy a large part of the working time for those who work with data.

Later in the chapter, you also read about how the first databases worked, and that perhaps the most important feature of a relational database is that it is searchable in a much more efficient way than previous databases. Finally, we concluded by examining how relationships, entities and attributes together constitute the logic in relational databases, and getting an overview of the rules for how they are created and interact.

When we want to structure and store data, databases of this type are often used, whether stored locally on our own machines, or in the cloud. And if you have also ensured that the data is well prepared—well, then you are well on your way to being able to utilise it.

Exactly how we proceed to use data to make better decisions and create value is the topic of the next and final chapter.