Collective search and recommendation in social media

Abstract

This PhD thesis proposal is focused on proposing solutions to the problem of collective search and recommendation in social media. User and data are two fundamental elements under social media environment. To cope with the semantic gap between social media data and semantic meaning, and the complexity of user intent and requirements, we propose to conduct research on three stages: (1) multimedia content analysis; (2) user understanding and (3)collective search and recommendation. We address the large-scale, multi-modal and heterogeneous characteristics of social media analysis by developing methodology from factor analysis, generative topic model and collaborative filtering. Progresses and advances along the three research lines have been presented, with future directions and open discussions concluded in the end.