Identification Of Students At Risk Of Low Performance By Combining Rule-based Models, Enhanced Machine Learning, And Knowledge Graph Techniques

الموضوعات

علوم تطبيقية وتكنولوجية

الملخص

Technologies and online learning platforms have changed the contemporary educational paradigm, giving institutions more alternatives in a complex and competitive environment. Online learning platforms, learning-based analytics, and data mining tools are increasingly complementing and replacing traditional education techniques. However, academic underachievement, graduation delays, and student dropouts remain common problems in educational institutions. One potential method of preventing these issues is by predicting student performance through the use of institution data and advanced technologies. However, to date, scholars have yet to develop a module that can accurately predict students’ academic achievement and commitment. This dissertation attempts to bridge that gap by presenting a framework that allows instructors to achieve four goals: (1) track and monitor the performance of each student on their course, (2) identify at-risk students during the earliest stages of the course progression (3), enhance the accuracy with which at-risk student performance is predicted, and (4) improve the accuracy of student ranking and development of personalized learning interventions. These goals are achieved via four objectives. Objective One proposes a rule-based strategy and risk factor flag to warn instructors about at-risk students. Objective Two classifies at-risk students using an explainable ML-based model and rule-based approach. It also offers remedial strategies for at-risk students at each checkpoint to address their weaknesses. Objective Three uses ML-based models, GCNs, and knowledge graphs to enhance the prediction results. Objective Four predicts students’ ranking using ML-based models and clustering-based KGEs with the aim of developing personalized learning interventions. It is anticipated that the solution presented in this dissertation will help educational institutions identify and analyze at-risk students on a course-by-course basis and, thereby, minimize course failure rates.

روابط وملفات

التعريف والنوع

رقم الوثيقة

f21ecefb-5b51-4fec-8413-34dc933eb6b2

رقم العقد

0

نوع الوسائط

Crawler

نوع المحتوى

الرسائل العلمية

صيغة المصدر

رسائل دكتوراة

نوع الملف

pdf text

أسماء الملفات

2325551_2.pdf

بيانات النشر

ألقاب المؤلفين

[{"name_ar":"Balqis Albreiki","title_ar":"اعداد","title_en":"Preparation"},{"name_ar":"Nazar Zaki","title_ar":"اشراف","title_en":"Supervision"}]

اللغة

English

المصدر والدورية

اسم المصدر

Identification Of Students At Risk Of Low Performance By Combining Rule-based Models, Enhanced Machine Learning, And Knowledge Graph Techniques

المحتوى والصفحات

عدد الصفحات

0

إشراف وإعداد

الإشراف

Nazar Zaki

الإعداد

Balqis Albreiki

الاقتباسات الببليوغرافية