论文信息 - Applying Visual User Interest Profiles for Recommendation and Personalisation

Applying Visual User Interest Profiles for Recommendation and Personalisation

We propose that a visual user interest profile can be generated from images associated with an individual. By employing deep learning, we extract a prototype visual user interest profile and use this as a source for subsequent recommendation and personalisation. We demonstrate this technique via a hotel booking system demonstrator, though we note that there are numerous potential applications.

[1] Ji Wan,et al. Deep Learning for Content-Based Image Retrieval: A Comprehensive Study , 2014, ACM Multimedia.

[2] Geoffrey E. Hinton,et al. ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[3] Rongrong Ji,et al. Weak attributes for large-scale image retrieval , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[4] S LewMichael,et al. Content-based multimedia information retrieval , 2006 .

[5] Trevor Darrell,et al. Caffe: Convolutional Architecture for Fast Feature Embedding , 2014, ACM Multimedia.

[6] Hermann Ney,et al. Features for image retrieval: an experimental comparison , 2008, Information Retrieval.

[7] Geoffrey E. Hinton,et al. Using very deep autoencoders for content-based image retrieval , 2011, ESANN.

[8] Nicu Sebe,et al. Content-based multimedia information retrieval: State of the art and challenges , 2006, TOMCCAP.

[9] Michael J. Pazzani,et al. Content-Based Recommendation Systems , 2007, The Adaptive Web.

[10] Victor S. Lempitsky,et al. Neural Codes for Image Retrieval , 2014, ECCV.

[11] Yang Song,et al. Learning Fine-Grained Image Similarity with Deep Ranking , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.