Towards social media as a data source for opportunistic sensor networking

The quality and diversity of available data sources has a large impact on the potential for sensor networks to support rich applications. The high cost and narrow focus of new sensor network deployments has led to a search for diverse, global data sources to support more varied sensor network applications. Social networks are culturally and geographically diverse, and consist of large amounts of rich data from users. This provides a unique opportunity for existing social networks to be leveraged as data sources. Using social media as a data source poses significant challenges. These include the large volume of available data, the associated difficulty in isolating relevant data sources and the lack of a universal data format for social networks. Integrating social and other data sources for use in sensor networking applications requires a cohesive framework, including data sourcing, collection, cleaning, integration, aggregation and querying techniques. While similar frameworks exist, they require long-term collection of all social media data for aggregation, requiring large infrastructure outlays. This paper presents a novel framework which is able to source social data, integrate it into a common format and perform querying operations without the high level of resource requirements of existing solutions. Framework components are fully extensible, allowing for the addition of new data sources as well as the extension of query functionality to support sensor networking applications. This framework provides a consistent, reliable querying interface to existing social media assets for use in sensor networking applications and experiments - without the cost or complexity of establishing new sensor network deployments