The Self-Adaptive Framework of Learning Object Based on Context

U-learning is coming to us. However, it is impossible for the learning object designed for traditional e-learning to be viewed appropriately without modification. In this paper, based on the existing research, a self-adaptive learning object framework is proposed for u-learning. This framework supplies different version of learning object to users according to their learning context which related to access device, network condition, and personal information. We design five versions for the learning object, original, HTML, Simple-HTML, WML, Simple-WML, which are corresponding to different learning context. Meanwhile, a self-adaptive learning object generation method is given based on the metadata K-node, and the conversion between versions of learning object is discussed.