Scenario Based e-Learning

Inspiration

Everything is going online. Traditional (e-)learning systems convert into forum, discussion like platforms with predefined files, pre-recorded videos etc. But problem based learning has a great opportunity to learn anything in any age range. This method is highly social learning event. Both social and theoretical skills are improved by dialog. In order to participate at the event, pre-study activities are improved intentionally. In every gathering, a teacher needs to be there for coordination or just keeping record of session.

What it does

We create/mimic all these features in to a web platform. Groups can be gather easily. Teachers can join in these live events. They can talk, send message or document. All of the content of scenario / problem sync through all group users. Teacher can hide or show themselves. This feature is useful because they can attend/watch more than one group. With this feature, group dynamics can be stable in every cases.

The main and non-trivial feature of our platform is its AI capacity. It can read and monitor all given data by groups objectively for grading and reporting purposes. English, Latin, Turkish languages are currently supported.

How I built it

We build it with PHP core (with HVMM), JS (async and native) for backend. MariaDB as a RDBMS database. Nginx for web server. libsodium for encrypting database values.

Challenges I ran into

Running a scalable, real-time communication web platform as easy as "Facebook" is a great challenge. Not everyone has the same network bandwidth. They always try to join/login from small screen devices and wants the same experience.

By using old but stable web language, PHP and using minimum server capacity is a real challenge that we accomplished. Groups can collaborate with real time audio or video (if needed), we do not collect or monitor their unrelated data. Only our AI monitors in order to understand each group member's contribution to the scenario/lesson. All the data that is collected are also visible to them.

Accomplishments that I'm proud of

Our NLP algorithm can work with medical case-studies. So that all around the world, a new case can be converted into a scenario and all the medical personnel can be educated by using the power of the connectivity. And all the gradings can be done by AI with absolute objectivity.

What We learned

Imaging and doing something from scratch is never easy. But it is absolutely worth the effort. We can see the need at this hard time to educate as much as people in order to help us.

Learn-Matrix