Mor Naaman

Social media data reflects people's attitudes, attention, and interests in a scale that was never available before. We use large-scale data mining and quantitative analysis to learn about society, language, communities, cities and more using social media data. We examine and leverage this information to create usable knowledge and applications, as well as develop and affirm theories of human communication and behavior.

Social media sites host an ever-increasing amount of user content captured or produced in association with real-world events. This project develops tools to find, organize, and present the social media content associated with events. Such tools can help ordinary users, as well as professionals (e.g., journalists) access event information from social media sources.

We use quantitative and qualitative methods to study how and why people use social media, and what are the social and personal implications of their activities.

At Yahoo!, we explored what type of information we can derive from Flickr geotagged images. We created Tag Maps / World Explorer as a new way to explore the world with geotagged images. We showed how to extract event and place semantics for Flickr tags, and suggested how one can create better search and representations for images of landmarks like the Golden Gate Bridge.

In my PhD, I have looked into organization of personal collections of photos. In particular, we tried to imagine the future of photo collection where photos are annotated with geographic location (imagine that!). Our early prototypes included PhotoCompas (organization) and LOCALE (sharing labels/tags between users).

At Yahoo!, we attempted to drive the future of mobile media creation and consumption with prototypes like ZoneTag and Zurfer. Both were context-aware, interactive mobile applications that were used by 1000s of users worldwide.