Grant: Collaborative Research and Development Grants

With the increasing popularity of social networks, microblogging has become widely accepted means of communication among people. Although the data in these microblogs are coarse and unstructured, useful information can be extracted from them. This extracted information can be used by applications for recom- mending products to users, performing large-scale data analytics and answering ad-hoc queries. We propose to develop a platform that allows the aforementioned applications to perform approximate search of content, at the same time dealing with high volumes and velocity of microblogs arriving in real time. The platform exploits content from microblogs generated in real time, and organizes it into a streaming graph. Continuous queries over the streaming graph will be useful for real-time monitoring of streaming microblogs. The graph will also be stored in a warehouse for offline social media analytics.