Social Learning Powered by NGraph

Karthik Deivasigamani
Noon
Published in
4 min readFeb 1, 2021
Image Credit : www.presencelearning.com

Context
As learning has moved online during these tough times, one major challenge faced by teachers and online learning platforms is how to keep the students engaged in the live classes. In this blog post, we will discuss social learning, students engagements in classroom and how we at Noon Academy are building a social learning graph to better engage students.

Social Learning
Wikipedia defines Social learning as “learning through the observation of other people’s behaviors”. As humans we tend to learn better when we observe and imitate somebody else’s actions and behavior. In the post-covid era, it has become increasingly evident to create a collaborative environment for students to interact & learn from not just their teachers but also their peers. At Noon Academy we had realized this long before covid and were working on building an open social learning platform. We incorporate social learning in our product through various interactive features (like groups, chat, interactive Q&A, breakouts) both within the classroom environment & outside. As we started adding 6M+ students to our platform building such features quickly required a strong platform that can help students discover the right teachers, groups & friends on the platform.

NGraph
NGraph is a Social Learning Graph which captures the interaction between Students, Teachers & Content on Noon Academy. The nodes in the graph could represent one of the following entities — student, teacher, content (course, subject, concepts) and the edges capture the relationship between them.
Before we dive deep let’s understand the relationship between students, teachers & content and why it’s important to capture them in the online world. Research shows that student engagement peaks when a student finds the content relevant to their interests, goals, and identities; feels that their relationship with their teacher is meaningful; and when the teacher has the appropriate levels of content and experience.

Image Credit: Where Student, Teacher, and Content Meet by Michael J. Corso, Matthew J. Bundick, Russell J. Quaglia and Dawn E. Haywood — 2013

NGraph works to not just capture the relationship between teacher and students, but also relationships among students. Through a series of A/B tests, we determined that students are more engaged when their friends are also in online sessions with them. Entities captured in our NGraph are :
1. Student
2. Teacher
3. Content ( subject, topic, academic concepts )
At Noon we are working on a concept knowledge graph by mining the table of contents of textbooks to build something like this —

Concept Knowledge Graph

We are able to capture the interactions between the above three entities in our NGraph as —
1. Student <-> Student : Our App allows students to send hi5’s to their peers for answering questions, exchange emoji’s, chat, upvote & comment on each other’s posts, etc. Also, our live sessions have breakouts where students are formed into smaller teams and allow for interactions (talk/chat) in a smaller team like setting. These interactions help us determine the affinity score between students.
2. Student <-> Teacher : Through the use of chat, voice, and emoji’s students are able to provide real-time feedback to teachers on their grasp of the concept being taught in the class. There are interactive Q&A, live competitions, and other feedback mechanisms to determine the student-teacher affinity score.
3. Student <-> Content : By breaking content into granular pedagogical concepts we are able to assess the relevance of them to a student based on their engagement scores within the groups or live sessions.
4. Teacher <-> Content : Noon Academy’s App allows the teacher to bring their own content or leverage existing content on the platform. Assessing the quality of the content, number of hours taught on a specific topic/subject, leaderboard scores we can determine the competence score of the teacher on these topics.

Below is a conceptual representation of the NGraph

Conceptual Representation Of NGraph

Application & Use Cases

Today at Noon Academy our featured groups and recommendation systems require a large amount of manual intervention to help students find the right teachers and learning groups. We believe that once fully built, NGraph can help us build the right discovery experience for our students & teachers. It will also help us create dynamic topic pages (collections) pertaining to subjects that a student of a certain grade is interested in.

Closing Remarks
Being an early stage startup we understand that building such a system will take time and require a lot of experimentation. Our approach to building this is going to be very iterative & incremental. In the coming months we will try to provide more details on the design choices & architecture of our NGraph. Watch this space for more updates!

We at Noon Academy believe in being bold, original, and ambitious. If you are passionate about education and love working on problems like these then email us at hrindia@noonacademy.com / ping me on LinkedIn.

--

--