Arguably the most difficult part of any project is coming up with a novel but executable idea. You can easily be trapped in the dataset rabbit hole.You spend hours looking through existing datasets and try to come up with new and interesting ideas. But here’s the problem with that approach: When you focus only on existing datasets — on Kaggle, Google Datasets, FiveThirtyEight — you limit your creativity to a small set of problems that this dataset was designed for.

At times, I enjoy exploring a dataset that sounds interesting to me. The one-millionth replication of a Kaggle prediction task may benefit my learning progress.** But as data scientists, we strive to create something new and unique and to generate useful insights.**

So how can you come up with a new idea? I combined insights from creativity research with my own experience. This resulted in 5 key questions including project idea examples.** These questions guide your idea generation process and enable you to leverage the full potential of your creativity. This will lead to new and unique data science project ideas**.

#data #creativity #research #data-science #development