With the rapid development of our society and non-stoppable renovation of information technology, the field of web-based learning is now drawing a lot of attentions from fields of researchers. And in a rapid speed people’s life is completely changed in every aspect. No exception the way of learning and receiving information are also greatly changed. Inevitably, we have encountered the self-directed learning at the digital age. No longer does one need to worry about no accessing to interested information which is expected to be on the internet whenever needed. And no longer does one have no ways of reaching to another people with similar learning interests who are expected to be connected by the cable wherever. Therefore, the two difficulties are greatly alleviated and the learning situation is greatly improved. Recently there is enormous number of information serving as learning resources on the internet.  Therefore, it has become possible to overcome the restrictions of time and place for people who are learning by themselves.

  However, like other new technologies, convenience always comes with consequences. There are three major difficulties when it comes to web-based self-directed learning and community-based learning. 1. It is difficult to locate suitable learning resources. 2. It is difficult to organize found learning resources. 3. It is difficult to pass learning achievements and get feedbacks from other members of a learning community. In this paper, I propose an approach which tackles the obstacles occur during self-directed learning on the internet. I designed a model, Multi-layer Map Model based on an ISO standard named Topic Maps. This model visualizes common learning behaviors on the web such as locating suitable learning resources, categorizing found resources and sharing them among community members. And then, according to the model proposed, a pilot system (A Resource Organization System for Web-based Self-directed/Community-based Learning) has been developed. I also propose an evaluation plan to test the effectiveness of this system to see whether or not potential users could improve the efficiency of their self-directed learning. After analyzing the experiment data, some positive conclusions have been drawn, several deficiencies have been detected, and most importantly many challenges are raised as future work of this research.