Παραγωγή κοινωνικών συστάσεων σε διαδικτυακή πλατφόρμα ηλεκτρονικής μάθησης με χρήση τεχνολογιών web 2.0
View/
Subject
Recommender systems (Information filtering) ; User interfaces (Computer systems) ; Διαδίκτυο (Internet) στην εκπαίδευση ; Information storage and retrieval systems -- Social aspectsAbstract
This master thesis refers to the construction of a system to provide personalized recommendations embedded in e-learning platform which deals with the subject "Application Development in Programming Environment" and is taught the last class in general high schools. Initially, it is presented the need for recommendation engines on Internet as explained in detail in this concept. Following are examples of websites that generate personalized recommendations for their guests. Then the categories of recommender systems’ methods are analyzed, as well as their advantages and disadvantages. Moreover some examples, as amazon, YouTube and google news, are analyzed that use algorithms from the category of collaborative filtering which finally we choose and develop in our e-learning website. In this category-collaborative filtering-, the similar users are calculated with the current (active) user of platform and recommend to him articles that have given high rank by specific users. A brief description of the platform is following which describes how the recommendation is produced, either partially personalized or not. Then is given a detailed description of the steps to modify the system, including the code of the algorithm that finally selected. In this way it was possible to the system to generate personalized recommendations to users. After that the evaluation of the system took place by students of B and C class sector IT of vocational school who have good knowledge of algorithmic structures. The evaluation of the system exported very positive results. The students were largely enthusiastic about the recommended individualized educational material and felt that they had a valuable assistant in their study. The construction of the platform was on the Content Management System JOOMLA 1.5.22 and producing recommendations using PHP, MySQL, HTML and extensions JOOMLA.