Remote Education Learner Classification via Behavioral Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The sudden shift to remote learning and exams posed challenges in ensuring fairness, reliability, and effectiveness, as well as identifying learners who may benefit from additional mentoring to improve their performance.
Innovation Solution
A computer-implemented method that collects data from actual, simulated, and exam environments to classify learners as cognitive or non-cognitive, applying a de-bias technique to identify those needing educational materials, and automatically providing tailored educational content to enhance their performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If remote exams are conducted without behavioral analysis, then exam administration is simple, but fairness and reliability deteriorate
Solution Approach 1:
The patent replaces manual invigilation and behavioral assessment with automated computer vision and machine learning systems. Cameras capture examinee behavior, and AI algorithms analyze facial expressions, head movements, and other behavioral patterns to detect cheating or engagement levels, substituting human judgment with automated mechanical systems.
Solution Approach 2:
The system introduces an intermediary behavioral analysis layer between the examinee and the exam results. This intermediary layer processes behavioral data through machine learning models to generate engagement scores and cheating detection, mediating the connection between raw exam data and final outcomes.
2Productivity
If all learners receive the same educational materials, then material distribution is efficient, but learning effectiveness for individual learners deteriorates
Solution Approach 1:
The patent applies local quality by providing different educational materials to different learners based on their individual behavioral patterns and performance data. The system analyzes each learner's engagement metrics, time spent on questions, and behavioral anomalies to customize material delivery, ensuring each learner receives appropriately tailored content rather than uniform distribution.
Solution Approach 2:
The system changes parameters of educational material delivery based on behavioral analysis. It adjusts material type, timing, and content based on detected engagement levels and performance patterns, transforming static material distribution into a dynamic, parameter-adjusted delivery system.
3Measurement precision
If behavioral data is collected continuously, then learner classification is accurate, but data processing load increases
Solution Approach 1:
The patent applies partial action by selectively processing behavioral data only when certain thresholds are exceeded or specific events occur during the exam. Rather than continuously analyzing all behavioral parameters, the system triggers detailed analysis only for significant behavioral changes or anomalies, reducing overall computational load while maintaining classification accuracy.
Data Source
AI summary
Provided are techniques for assisting remote education learners. Data is collected for an actual base environment, a simulated base environment, and an exam environment for a learner taking an exam. The collected data is used to generate a first behavior pattern for the learner in the simulated base environment and a second behavior pattern for the learner in the exam environment. In response to the second behavior pattern deviating beyond a threshold from the first behavior pattern, a classification of non-cognitive learners class is determined for the learner. A de-bias technique is applied to the classification to generate a final classification. In response to the final classification being the non-cognitive learners class, education material for a subject covered in the exam is selected, and the educational material is played on a remote learner computer of the learner.