The problem SMARTLearn solves

Our goal is to provide adaptive and personalised learning to students. A student's weak areas are identified and relevant and necessary resources are provided for improvement. We are focusing on students from KG to 12th standard over multiple boards. Videos covering all the syllabus chapters and subtopics are provided. A quiz also gets generated using transformer models on each video. Based on some video watching metrics and the quiz score, the weak areas of the student are recognized, and relevant videos are recommended using Q-learning, a reinforcement learning algorithm. PDF resources are also provided topicwise. A dashboard for the student is also given.

Challenges we ran into

Database: Initially we found a OULAD dataset, but it didnt have required parameters. So we prepared a synthetic dataset on necessary trends for the purpose.