文章目录
本篇论文是关于线上学习的推荐算法,算法的一大优势在于它是 online 的,具有比较好的适应性和多样性,同时它基于的不是用户与用户之间的关系,而是去探索用户与学习目标,以及学习目标之间的关系。在 online learning 这一块,这样的方法被认为是比较科学的。
Abstract
E-learning remmondation system aims to provide learners with learning objects (LO) that fit their learning goal and hence assist in learning. Traditional content-based (CB) remmonder system are mostly based on the matching rules between learners and learning rules. Such approaches are limited when learners change constantly and have low adaptability and diversity. In this paper, in order to improve the adaptability and diversity of recommonder system, we incorporate LO-oriented recommondation mechanism and propose a LO self-organization based recommendation approach, which is able to response according to the relation among LOs. The proposed approach works as follows: firstly, LOs are modeled as intelligent entity with multiple attributes th