LeetPath recommends LeetCode problems based on your past interactions, ensuring each recommendation is tailored to your current skill level.
LeetPath considers the difficulty levels of problems you've solved, providing a balance of challenge and progression.
As you solve more problems, the system adapts and ensures the recommendations always reflect your evolving skills.
A graph-based recommendation engine connects questions dynamically, improving recommendations over time.
Questions are modeled as nodes in a graph structure, with relationships between them reflecting content similarity.
We identify latent topics to match questions to your specific skill gaps for improved relevance.
We use belief propagation in Markov Random Fields to refine recommendations based on probabilities.