Skip to content

dudupopkhadze/knowledge-graph

 
 

Repository files navigation

Vitsi AI

Vitsi AI is an educational startup that is built on the concept of representing knowledge as a graph, where nodes represent concepts or topics, and edges represent relationships between them. The platform aims to provide a personalized learning experience to learners of all ages, from kindergarten to adulthood.

Problems with current education system

  1. Students do not know what they know and what they don't. They have gaps in there knowledge they are not aware of. And it's pretty hard to learn new things when you do not quite understand some of the prerequisites

  2. Students don't know what they want to learn. For example, there are lot's of students who think that they want to become a software engineer(because they heard it pays a lot), but when they start coding they start to think that it's not for them or it's not something they enjoy.

  3. Even if they have the goal and know what they want to learn, as a rule, there is no clear path to it.

    1. There is a formal education path, when you go to the university and you trust that what they teach you is the relevant knowledge. Of course in some sense any knowledge is relevant, but there is a notion of an "alternative cost". In universities students are forced to learn courses that they do not like and may never use. The courses themselves are outdated notion as well, because some might teach the relevant knowledge, but not everything will be relevant for everyone who is taking the course.
    2. There is also an informal education path, where we have lots of boot-camps, MOOCs, etc. They too mainly mimic the university courses and provide even less clear path towards the goal, and less trust.
  4. Because of reasons listed above and many others there are lot's of students who start but do not finish courses. And they finish learning. Which is ironic because you can finish a course but never the learning.

Solution we propose

The knowledge in and of itself is a direct acyclic Graph. The nodes of the graph are tiny bits of knowledge that is easy to understand and learn and edges are the prerequisites of this knowledge. So basically if you know all the previous(prerequisite) nodes you can learn the new one. And all the knowledge of the humankind can be presented like this.

With the knowledge graph like this built, we can clearly see everything there is that could be learned. We can see what we already know and if we want to learn something what's the shortest path we might want to take to get there. Knowledge graph does not offer courses in a specific field that you need to complete, but the experience is based on the concept of lifelong learning, where small “knowledges” acquired on the regular basis lead you to mastering in a comprehensive and extensive manner.

In the Knowledge Graph each Node consists of name, description, resources(such as videos and texts) and tests. If student passes the tests they "know" the node and can learn nodes leading from there. Users can upload their own content for each node to help each other.

There is a short video mode of learning similar to youtube shorts, facebook reels or tiktok, where students discover different fields they might be interested in, by watching the videos(honoring the prerequisite graph).

Run Knowledge Graph

Follow the commands here

Interested in contributing?

Please see CONTRIBUTING.md. Also join the Discord server to join discussions about contributions, development, and more.

About

Explore the world's knowledge, curated just for you

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • TypeScript 52.9%
  • Go 45.9%
  • HTML 0.4%
  • Shell 0.4%
  • Makefile 0.2%
  • JavaScript 0.2%