Learning Objects Automatic Semantic Annotation by Learner Relevance Feedback

To search learning objects, in particular, multimedia resources quickly in an e-learning environment, additional semantic information should be attached to the objects. Attaching and using this semantic information refers to three respects: semantic representation model, semantic information building and semantic search techniques. In this paper, we introduce an associated semantic network as the semantic representation model; use semantic keywords, a linguistic ontology in semantic similarity calculation and use learner relevance feedback to complete automatic semantic annotation. After several iterations of learner relevance feedback, semantic network is enriched automatically. In addition, semantic seeds and semantic loners are employed especially to speed up the growth of semantic network and to get a balance annotation.