The effect of current WEB-based teaching is dissatisfactory. Learners tend to indulge in the sea of learning resources and they are unable to find suitable resources, information due to the overloading. As a result, it is not uncommon to see that learning anxiety happens and learners become "lonely after the excitement" in this teaching method. Therefore, the current situation of learners in WEB-based learning gives some cause for the concern.Many educational technologists try to draw on e-commerce recommender technology, by tapping into learning features to help learners find "individual" learning resources, which are proved to be effective to some extent. In this way, however, the traditional way of teaching "teacher cramming education" is replaced by the network-based teaching "resource cramming education." The author of this thesis believes that this is a commercialized act, which is lack of in-depth understanding of the needs of network learners.Based on relevant literature and the analysis of difference between traditional teaching environment and network teaching environment, the author of this thesis pointed out that the key lies in the learning environment. The advantage of network teaching is also its biggest drawback, which directly reflected in the lack of an effective collaboration among network learners, and the effect of "role models" In network teaching environment, "peers", "experts", "learning partners" are extremely valuable learning resources, which are the pulling power and impetus for the learners. Recommending suitable human-learning resource will be critical factor for success.This thesis is based on this concept and WEB2.0 social tag technology to explore learning features, build learner modeling, and then find similar learning partners and conduct a recommendation for them.Social tag technology is appealing delicacy of WEB2.0 era. Learners manage personal resources through social tags, which is an individual behavior and reflects learners'understanding and awareness of resources and their interests and preferences for resources, endowing the tags with the potential and value of exploring learners' behavioral features. |