Disparity-preserved Deep Cross-platform Association for Cross-platform Video Recommendation

Cross-platform recommendation aims to improve the recommendation accuracy through aligning or fusing informationfrom different platforms together. Existing cross-platform recommendation approaches assume all cross-platform information to be consistent with each other and can be aligned. However, there remain two challenges unsolved by previous works on cross-platform video recommendation: i) there exist inconsistencies in cross-platform association due to the platform-specific disparity, and ii) data from distinct platforms may share different semantic granularities. In this paper, we propose a cross-platform association model for cross-platform video recommendation, i.e., Disparity-preserved Deep Cross-platform Association (DCA), taking platform-specific disparity into consideration to address these challenges. The proposed DCA model employs a partially-connected multi-modal autoencoder capable of explicitly capturing and preserving platform-specific information in deep latent representation, as well as utilizes nonlinear mapping functions to handle granularity differences from various platforms. We then present a cross-platform video recommendation approach based on the proposed DCA model, which automatically concentrates on valuable cross-platform information from a comprehensive semantic level. Extensive experiments for our cross-platform recommendation framework on Twitter and YouTube datasets demonstrate that the proposed DCA model can significantly outperform existing cross-platform recommendation methods in terms of various evaluation metrics.