Design And Implementation Of Personalized Teaching Resources Recommendation System

With the information technology developing rapidly, the Internet has been the platform for information sharing and businesses are taking advantage of the Internet to accelerate their informatization process to meet the demands for access to and sharing of information. However the concomitant is the information explosion problem as users sharing too much information on the Internet. They might have difficulty in the flood of information from the Internet to find what they want or to express their needs, so search engine and recommendation engine appeared.In the education field, the informatization process is also developing rapidly and teaching assistant system has been used to a lot of teaching practice. As a part of teaching, teaching assistant system can help teachers to share educational resources to students and help students to communicate with other students. This system can be seen as a complement to teaching on the course and students can learn through this system after class according to their own needs independently.This thesis designs and implements a personalized recommendation system for teaching assistant system which recommends educational resources to users according to their browsing history, browsing habits and preferences. This recommendation system uses a hybrid recommendation model based on artificial neural network consists of four sub-models which are recommendation models based on resource contents, item-based collaborative filtering, user-based collaborative filtering and artificial neural network and the output of the former three sub-models is input of the fourth sub-model which will records users’preferences and habits for personalized recommendation. Besides, this recommendation system uses a two-step training method to solve the user cold-start problem. This recommendation system is a good solution for personalized teaching resources recommendation.