近年来,人工智能在教育领域的发展取得了重大突破,其中多模态数据融合与智能教育相结合的应用愈加广泛,其可能引发的伦理风险也备受社会关注.本文详细阐述了多模态数据融合下智能教育的发展现状,深入剖析了智能教育可能引发的伦理问题,最后提出解决现有伦理困境的政策措施,以期推动智能教育向善发展,更好地服务于人类和社会.

The moderate level of collection development is an important limitation to the library collection development.Pareto principle,collocation of core resources theory,collection structure theory,long tail theory and the goal theory of "integrate disciplines puzzle" have great effects on determining the moderate level of library information collection development.These theories emphasize distinctive construction of key collections and general collections,or emphasize the integrity of the collection development.The goal of "integrate disciplines puzzle" requires two divisions of collection structure,which contains "coconstruction district" and "self-construction district".The collection development in "co-construction district" must strive to reach the integrate level to meet the need of co-construction and sharing,and form the "integrate disciplines puzzle" with the resources in "co-construction district" of the other co-construction institutions.While the collection development in "self-construction district" can determine the moderate level according to the reality of the institutions and the other theories.