Nearest Acquaintance Modeling: A Multi-Objective Approach for Cohort Formation and Instructional Material Design Using R-Tree Data Structures

One important aspect of enhancing the student learning experience is aligning the instructional materials with their exact preferences. While customizing the instructional material to suit the needs of each individual student presents a significant challenge, most educators aim to align the materials with a cohort of students who have similar needs. This paper introduces Nearest Acquaintance Modeling (NAM), a multi-objective approach for creating student cohorts and designing instructional materials using pre-assessment scores and learning styles. The method utilizes the exceptional functionality of R-Tree data structures to efficiently manage multidimensional student data, ensuring optimal cohort formation.To better understand the learning profiles of each student, we collected pre-assessment scores and learning style preferences using Allinson and Hayes' Cognitive Styles Index (CSI) and Kolb's Learning Style Inventory. By leveraging R-Trees mappings, we identified student cohorts with similar learning needs while promoting diversity. Customized instructional materials were then provided to each group, which significantly improved students engagement and knowledge retention.The results presented on a case study with 100 first-year engineering students indicate a marked increase in post-assessment scores and significant satisfaction from students. The NAM approach, therefore, makes an in-depth case presentation for a larger study on education as an optimizable variable utilizing a data-driven approach, which leads to further research on the personalization of learning.