Abstract: In the Web environment, resources have rich contents, various forms, different description standards as well as discrete organizational structures. Problems, such as the lack of semantics and domain knowledge and artificial workload, limit them in the personalized resource recommendation and the obtaining application. This paper, based on the relationship between the concept of activated diffusion and the thought of layered resource semantic identification, establishes the relational concept annotation model (RCAM）. RCAM simulates human memory activation spreading process and considers memory enhancement and forgetting mechanism to achieve the more dynamic and personalized related organizations of Web resources. RCAM regards the concept of a relationship as semantic elements, and the fragment relationship concept set as the semantic scheme. With the flexible identification granul-arity and relatively completed sematic logic, RCAM provides new research ideas to Web education resource identification. Experiment shows that RCAM can provide more background knowledge, adapt to dynamic resource organization with different learning scenarios and learning personality, and it aims at the open Web resources. As a result, RCAM has better universality and extensibility.

Key words: Web resources annotation, relational concept, activation and spreading, relational concept annotation model (RCAM)

摘要： Web环境下资源内容丰富，形式多样，描述标准不一，组织结构离散，既有标识方法存在语义单一，缺乏领域知识背景或人工参与度大等问题，限制了它们在个性化资源推荐和获取中的应用效果。基于关系概念的激活扩散以及资源分层语义标识的思想，面向Web资源构建了关系概念语义标识模型（relational concept annotation model，RCAM）。RCAM模拟人类记忆激活扩散过程，考虑记忆的加强与遗忘机制，由此实现资源更为动态、个性化的关联组织。RCAM中资源标识以关系概念作为语义要素，以片段关系概念集为语义模式，标识粒度灵活，语义逻辑相对完整，为Web资源标识组织提供了新的研究思路。实验表明，RCAM能提供更多的领域知识背景，可适应不同学习情景、学习个性下的资源动态组织，并且针对开放的Web资源，具有更好的通用性、扩展性。

关键词: 网络资源标识, 关系概念, 激活扩散, 关系概念标识模型（RCAM）