Towards an effective recommendation algorithm for education domain

Knowledge is the key element that guarantees the existence and survival of the human race. Throughout the history of mankind, we see rapid evolution in the way we live, work and play. However, the only aspect of our life that changes slowly is education. Even since the last 50 years after technology...

Full description

Bibliographic Details
Main Author: Choun Nov
Other Authors: School of Computer Engineering
Format: Final Year Project (FYP)
Language:English
Published: 2014
Subjects:
Online Access:http://hdl.handle.net/10356/58926
Description
Summary:Knowledge is the key element that guarantees the existence and survival of the human race. Throughout the history of mankind, we see rapid evolution in the way we live, work and play. However, the only aspect of our life that changes slowly is education. Even since the last 50 years after technology boom and the invention of the Internet, we hardly change the way we teach our children. The so-called online elearning system and many fancy electronic devices that claim to change our education system are just using the same teaching methods but in a shiner way. The purpose of this project is to make enhancement to an existing elearning system, which was developed in the past by senior NTU students. The system was designed to revolutionize our education system teaching method by introducing more interaction and collaboration between students and professors. The original system was able to predict students grade and recommends materials to students. There was also a web crawler that used to crawl for material, and a notification module that sent email to students. The system was written in JSP and HTML with Apache Hbase and Hadoop as its backend storage. For the enhancement works on the system, the improvements include removal of prediction module and web crawler, recommendation module modification and change of database from HBase to MySQL. New modules include introducing path planner for the course so that professors can cater specific assessments to a diverse groups of students. Student grouping algorithm for assisting professors to assign students into groups based on their assessments’ results. Group wiki and forum discussion for students to exchange ideas and collaborate. Graphical representation of students' assessments results for professors to analyze student’s performance. The enhancement works are 100 percent complete and tested, and received commendation for great innovations and user friendliness. For the next round of system improvement, perhaps, modules of existing elearning systems such as announcement can be developed and integrated into the current system.