Jupyter Notebooks: High Interaction/High Scale Learning – by Bill Brandon

At The eLearning Guild’s DevLearn 2018 Conference & Expo, I sat in the back of Karen Hebert-Maccaro’s presentation, “The Technologies Enabling New Approaches to Learning,” and had an epiphany moment. In her session, Hebert-Maccaro introduced two concepts: performance adjacent learning and high interaction/high scale experiences. Among other technologies, performance adjacent learning uses voice interfaces such as Alexa. She exemplified the other concept with the Jupyter Notebook, along with virtual, augmented, and mixed reality applications.

This article introduces the Jupyter Notebook and includes links to sources where you can learn more about it, including how to make Jupyter Notebooks.

Jupyter Notebooks are already familiar to those in education, data science, and a few other fields, but the approach has many uses relevant to learning—most as yet undiscovered or unrealized within L&D. There will be more articles to come about these interactive asynchronous environments. Think of this as an appetizer.

The shape of things to come

Two years ago, an eLearning Guild research report asked 13 thought leaders if instructional design was a dying art. In a Learning Solutions article, Chad Udell observed, “I sense a great deal of fear, uncertainty, and doubt coming from a wide swath of the ID community, unfortunately. People have shifted in how, when, and why they access information to do their jobs better, more safely, and more productively. The move to the always-on, ubiquitously connected workforce has tilted the power equation firmly in favor of the learner, and the ID community is largely still reeling in a Who Moved My Cheese?-type of fashion.”

Udell’s observation was prophetic. Today, we can more easily see the disruption that is coming to “the way we have always done it.” Consider the emergence of learning engineering as a cross-disciplinary approach to learning. Many L&D practitioners have a growing understanding that in order to remain relevant to business and the needs of employees—both of which are concerned about improving performance in the face of constant change—we must think in terms of supporting self-directed learning more than teaching. Meeting all moments of learning need and learner proficiency requires more than what most learning and development work involves today; on their own, traditional training, education, and performance support do not respond adequately across the Five Moments of Need.

What is coming next is the move to learning experiences that employees can access when needed, where needed, without requiring in-person assistance from an instructor, coach, facilitator, or expert. These experiences go beyond building canonical knowledge. Designing these experiences will become a much bigger part of our job in the very near future. Fortunately, technology is here to help. The technology may involve virtual reality, augmented reality, immersive simulations, and documents or web applications such as the Jupyter Notebook, and similar interactive asynchronous environments.

What is a Jupyter Notebook?

The Project Jupyter website defines it this way: “The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text.”

In a follow-up interview for this article, Hebert-Maccaro said, “The real power in Jupyter Notebooks as both a learning and a practitioner tool may be in the fact that it allows individuals to engage in an experiential environment to do analysis that supports a contextualized view of data and information. It allows for a rich, complex, and nuanced analysis. As such, it requires interpretation and contextualization of the result before application to a decision or action. It is the opposite of a one-size-fits-all approach to thinking and learning.”

There are thousands of these Notebooks already in use. As Hebert-Maccaro pointed out in her presentation:

  • Data scientists use them; in a Kaggle survey about 8,000 data scientists recognized the Jupyter Notebook as the fourth most-used data science tool for work.
  • Amazon SageMaker is a fully-managed machine learning service in which data scientists and developers can quickly and easily build and train machine learning models, and then directly deploy them into a production-ready hosted environment. It provides an integrated Jupyter-authoring Notebook instance for easy access to data sources for exploration and analysis, without requiring server management.
  • Students learn to code in Python and other computer languages through Jupyter Notebooks.
  • At UC Berkeley, Data 8 is the first undergraduate degree program in data science, enrolling 600 students in each course—based on Jupyter Notebooks.
  • O’Reilly Media provides books, instructor training, and other resources (Orioles) using the Atlas platform to make Jupyter Notebooks a first-class authoring environment for its publishing program.

Similar concepts include Wolfram Notebooks, Azure Notebooks, and the Spark Notebook. Each of these has its own specialized purpose with particular features that facilitate use and management, but all are interactive, asynchronous environments

How can these Notebooks support learning? While a Notebook somewhat resembles a textbook in terms of layout and content, it is the live code, equations, and visualizations that make them most useful to self-directed learners. Users can modify the code and run it as part of the learning experience. They can change the inputs and observe or study the changes in the visualizations. Users can practice and learn from their mistakes.

Try out a Jupyter Notebook here

To give you some experience with self-directed learning involving building a Jupyter Notebook, try this book. Open the link and then open the link named “A demo of the hosted textbook.”

If you are more comfortable simply experiencing a course built with Jupyter, try this textbook from the UC Berkeley Data 100 course, “Principles and Techniques of Data Science“.

In the comments section below, please let me know if you have any questions, or what you would like to see in the future articles on Jupyter Notebooks (or other Notebook environments).

Read more…

Leave a Reply