LeetPath recommends LeetCode problems based on your past interactions, question similarity, and topics you're most likely to excel at, ensuring each recommendation is tailored to your current skill level.
LeetPath considers the difficulty levels of problems you've solved and provides recommendations that balance challenge and progression, making sure you're continuously improving.
As you solve more problems, the system continuously learns and adapts, ensuring the recommendations always reflect your evolving skill level and learning trajectory.
A graph-based recommendation engine builds connections between problems based on content similarity and topic overlap. This results in dynamic, interrelated recommendations that improve over time.
A graph structure models questions as nodes, with relationships like content similarity and topic overlap as edges. This allows for smarter recommendations based on related questions.
Latent topics within questions are identified, helping to improve recommendation relevance by matching questions to your specific skill gaps.
A Markov Random Field (MRF) is used for belief propagation, refining recommendations through joint probabilities to ensure they match your skill level and learning preferences.