Content Based Video Recommendation for Multimedia Big Data

Abstract

Online social networks have been massively growing multimedia services and offer huge amount of video contents. Because of that users face some problem when they search for particular video. From that huge amount social media user delay in what they want. For that purpose personalized video recommendation system created for the users. This recommendation system solves the problem user when they search for particular video. But no one has consider both the privacy of user context (e.g. social status, ages, profession and hobbies) and video service providers, which are extremely sensitive and signicant in commercial value.To handle these issues, derive a cloud-assisted differentially private video recommendation system based on distributed online learning. To achieve this, service vendors architecture is modeled as distributed cooperative learners, recommending videos according to users context, with simultaneously adapting the video-selection strategy based on user-click.The sparsity and heterogeneity of massive social media data may reduce the performance loss of the system. Algorithms outperform existing method and keep delicate balance between computing accuracy and privacy preserved. And also increases efciency and save time when user search for particular video in social media.