eLORM: Learning Object Relationship Mining based Repository

The development of learning objects (LO) has brought significant promise for creating highly personalized learning programs and it has resulted in the huge amount of various LOs all over the world. Learning object repository is regarded as a good median to manage the learning resources such as storing, indexing, retrieval and so on. However, due to the LOs´ distribution and diversity, learners usually lack sufficient awareness and comprehension to make effective selection among the large numbers of available learning resources. It is difficult for the learners to properly integrate information to address their immediate learning need, for example, the related learning resources, or prerequisite learning domains, etc. In this paper, we put forward the idea of mining LO relation based on the learners´ usage information in our LO repository (called as eLORM). By analyzing the LO relation patterns from their learning domain, granularity, categories, etc., the correlated various learning objects are able to be recommended to the learners. Based on the definitions of the LO relation patterns such as association pattern and sequence pattern, we apply the web usage mining methodologies in our LO relation patterns´ model. We also develop a LO repository system based on our LO-relationship-mining approach.