I am Preparing the Ultimate Data Science Roadmap

Boy: Do not try and bend the spoon. That's impossible. Instead only try to realize the truth.
Neo: What truth?
Boy: There is no spoon.
Neo: There is no spoon?
Boy: Then you'll see that it is not the spoon that bends, it is only yourself.

In the Data Science Facebook Group, I often see new students asking 10+ years experienced professionals about sharing a roadmap to get into the field. It is understandable that many new learners will be overwhelmed by the complexity of the field. To worsen it more, there is so much advice on the internet, thousands of free and paid courses, textbooks, tutorials - it is normal to feel confused.

When I was asked this question n-th time, I just answered "There is no roadmap"

This answer may sound fun if you know how to deliver the punchline - but doesn't satiate the intellectual need of a young soul who sincerely wants to learn. So I opened my Obsidian valut I use only for preparing teaching material, and looked at how the pages were connected to each other, and how do they correspond to different lectures, tutorials, and courses I deliver.

Neo: What truth?
Boy: There is no spoon.
Croc. Dundee: This is a spoon.
Magritte: This is not a spoon.
The Tick: SPOOOOOOOOOOON!!!

I realized that the map of Data Science or any other complex field of expertise is more like Grand Theft Auto, not like Road Rash. Holding hands and preparing a straightforward roadmap wouldn't do justice to the curiousity of a learner. There are many attempts on preparing the roadmap - most of them were created by tutorial websites, coaching institutes, and bootcamps. They represent the whole Data Science journey as a series of courses one could take, again, leading to the same structure of a roadmap. Another Metro Routemap like chart shows a similar series of topics/chapters/courses presented in a more fun way, rather than something that would comprehensive nature of the field.

But Why Do We Need A Better Roadmap?

I agree - someone studying in a university or buying a course from a Bootcamp would like to have a clear idea about the sequence of topics they will be learning. But when you place Python before Statistics, or Visualizations before Machine Learning, you are influencing the young mind into thinking that these can not be studied out of order. The fancy roadmaps and course guides available on the internet (some people even sell you eBooks called something like "How to Learn Data Science" for real money. I grind my teeth when I see people who are students themselves preparing and selling these guides with no clear pedagogical approach and limited thoughts on the field. Somebody has to do it.

In next few weeks, I will be working on collecting resources, course plans, ontologies and concept maps to systematically find the building blocks of this field in a different granularity. Some of the building blocks will be connected to each other weakly, while some will be connected strongly - thus indicating pre-requisites, sub-topics, or parallel topics. I will collect freely available courses, tutorials, book chapters and assign them to each topic.

How Long Will This Take?

Honestly, I don't know. This is a large project. If you are learning Data Science or you are a teacher, feel free to get connected and ask me for a copy of my work-in-progress. If you would like to contribute, you are welcome to join me and my collaborators who will be working on this.