356 Building a Learning Path Recommender - Manual Construction of Knowledge Graphs in Python

In this tutorial, we dive into the construction of a knowledge graph for educational content using a manual, expert-driven approach. The example code focuses on building a learning path recommender for videos from DigitalSreeni's YouTube channel, with topics covering:

Core Python programming concepts
Financial analysis and algorithmic trading
Bioimage analysis and microscopy

Rather than relying on automated discovery, this approach explicitly defines the nodes (topics) and edges (relationships), leveraging domain expertise to create accurate and meaningful learning paths. Each topic is annotated with metadata such as difficulty level, keywords, domain classification, and detailed descriptions. Relationships between topics are also manually weighted, indicating prerequisites, content similarity, and learning path progression.

The knowledge graph is visualized using interactive (Pyvis) and static (Matplotlib) methods, offering clear insights into how the topics connect. The graph data is stored in a SQLite database for persistence and easy querying.

While this manual construction method ensures high precision, it can be time-consuming to maintain, especially for larger content libraries. In future examples, we’ll explore AI-based approaches that can automate relationship discovery and adapt to new content more efficiently.

By the end of this tutorial, you’ll gain practical knowledge of building a knowledge graph to organize and recommend learning paths based on educational content.

Link to the code: https://github.com/bnsreenu/python_fo...