Recommender System

Every learner is unique, not only in terms of their background and Cognitive abilities, but also in their preferences and motivations in various subjects and fields.

Furthermore, an individual's level of motivation and engagement can fluctuate over time.

For Adaptive Learning protocols to be most effective, a Recommender System is essential.

At EdCortex, we prioritize assisting EdTech companies in implementing a recommender system tailored to their specific requirements and capabilities.

Some examples of our algorithms include Multimodal Embedding and Classification, Approximate Nearest Neighbors, Bayesian Personalized Ranking, and Matrix Factorization.

To this end, we introduce a novel and scalable set of methods and services that are being developed at EdCortex, allowing the online learning providers to comprehensively measure the learning preferences and Cognitive traits of their learners through Questionnaires, Cognitive Tasks and Performance measures. This metadata is then used to securely classify each learner into a distinct profile, characterized by learner's preferences, specific learning needs and Cognitive skills. 

Each profile is then accompanied by a dedicated learning path, instructed by the latest findings in the Science of Learning, that would best maximize the learning KPIs of the learner, modulating for the learner the notifications, feedback messaging, error-rates, content sentiment characteristics, etc.

Here you can find more information on EdCortex Profiling methodology.