Multimodal Grand Challenge Prize Winner: Lexing Xie
Congratulations go to Dr Lexing Xie and her collaborators for winning first prize in the Multimodal Grand Challenge category at the ACM Multimedia conference held in Nara, Japan.
Their winning entry was a multimedia magazine presenting trending topics and news events via statistical microblogs analysis.
In a world of practically instant communication, many events are first reported in social media. Increasing amounts of rich-media content (pictures and text) are associated with such posts, making them more credible and attractive. Dr Lexing Xie and her collaborators designed a rich-media analysis system to address the important challenge of sensing and exploring events from social media in real-time.
The system includes a novel bi-lateral correspondence topic model to extract representative content and meaningful facets (defined as related to the who, when, where, what, why and how of event content, consistent with the 5W1H principle in news reporting) about events of time. It also includes a digital magazine that anchors user interactions with event facets.
Lexing and the team presented a multimedia event analysis and exploration system that has three main components: abundant data sources of millions of media-rich microblogs, bilateral correspondence model for modelling the relationship and trend between text and image, and faceted event browsing in a magazine design. To the best of their knowledge, this is the first system that analyses event-centric rich media with principled statistical models, and constructs event facets for interactive social media events exploration.
Their system won first prize in the Multimodal Grand Challenge category at the premier worldwide multimedia conference ACM Multimedia 2012, a key world event displaying scientific achievements and innovative industrial products in the multimedia field which was held earlier this month in Nara, Japan.
Congratulations, Lexing!
