Nthexam ML and memory based recommendation for student study pattern
Turning data into action for products beyond analytics
We aim at providing the best and concise quality content to student and content users. Nthexam personalized book is based on our rigorous data research and machine learning. Each author, teacher and student feedback is the driving force to make the best of the data useful for personalization of student content. With our personalization, each student is connected to millions of authors and vice versa.
We work on progressive growth of student
Each student has its own pace and learning curve and a single teacher cannot concentrate on all the student equally, this is where our AI is working to provide a progressive growth strategy for each student
Making the one to one learning using technology
Each student has its own pace and learning curve and a single teacher cannot concentrate on all the student equally, this is where our AI is working to provide a progressive growth strategy for each student
So how nthexam provide personalisation?
Factors we take into account
- Marks secured in different categories, sub categories(subject), sub sub categories(topics)
- Difficulty levels
- Time taken
- Memory retention
- Creating cohorts
- Behaviour pattern of attempts
- Accuracy index
- Weighted pattern for each type of mock test, rankers test, personal and general assignment, self created, practice sets, nthexam recommendations, pro-metric tests.
- Relative AND Absolute performance
We have a pool of performance data at our backend ie, database.
The performance calculator API, personalised recommendation API, score and percentile
scaling API, Memory pattern analyser API, are used to generate the recommendation. Each
recommendation is generated with an expectation score and which on each attempt is
updated.
The forgetting curve:
We follow a half life pattern for each category, subject and topic for each type of difficulty levels of questions.
We take various attempt at different levels and from different pattern and it needs to be scaled on a system
This is first version of the institute and student recommendation platform for the IIT, NEET and GATE exams. We are working hard to provide much accurate and precise recommendations.
