In this paper,a novel image tag recommendation framework is developed by fusing the deep multimodal feature representation and cross-modal correlation mining,which enables the most appropriate and relevant tags to be presented on the image and facilitates more accurate image retrieval.Such an image tag recommendation pattern can be modeled as an inter-related correlation distribution over deep multimodal visual and semantic representations of images and tags,in which the most important is to create more effective cross-modal correlation and measure what degree they are related.Our experiments on a large number of public data have obtained very positive results.