【摘要】 交叉学科具有比单一学科复杂且知识关联度更强的知识结构，基于传统单一学科设计的资源推荐算法无法直接推广到交叉学科。为解决此问题，提出了一种融合知识图谱与协同过滤的数字化资源推荐算法，以帮助跨学科学习者了解学科间的潜在联系并实现知识的跨领域迁移与整合。通过知识图谱可视化学科间的知识网络，计算数字化资源间的语义相似性；利用协同过滤的方法计算课程相似度，通过融合层给出推荐结果。基于国家高等教育智慧教育平台资源数据的实验结果表明，该算法相比传统的推荐方法，在精确度、召回率等指标上均有较好的表现，验证了所提方法的有效性，缓解了数据冷启动问题，可以为跨学科学习者提供更为合理的数字化学习资源规划及拓展。更多还原

【Abstract】 Interdisciplinary fields possess more complex knowledge structures and stronger interconnections than single disciplines, making traditional resource recommendation algorithms designed for individual disciplines unsuitable for direct application. To address this challenge, an interdisciplinary digital resource recommendation algorithm integrating knowledge graphs and collaborative filtering was proposed. Potential relationships between disciplines were identified by the algorithm, facilitating cross-domain knowledge transfer and integration. The knowledge network between disciplines was visualized through a knowledge graph, and the semantic similarity between digital resources was calculated. Course similarity was computed using a collaborative filtering approach, and recommendation results were generated through a fusion layer. Experimental results based on data from the smart higher education of China platform demonstrate that the proposed algorithm achieves higher precision and recall compared with traditional methods, confirming its effectiveness. It alleviates the cold-start problem and provides more rational digital learning resource planning and expansion for interdisciplinary learners.更多还原