A SERVICES FRAMEWORK FOR USING THE SOCIAL WEB AS SENSOR WEB: IDENTIFYING EVENTS IN REAL TIME USING SPATIAL, TEMPORAL AND SEMANTIC REFERENCES (SUSS).

In this paper we discuss the design of a system that uses the social networks to detect events using the social networks' users as sensors that report information. The system is using text mining and machine learning algorithms to cluster tweets and social media posts and then use spatial and temporal metadata to identify events, as well semantic information to identify the sematic relations among the different concepts involved. We describe the steps to follow and the algorithms we implement in the system in order to achieve this. At the end we also discuss the overall user functionality offered to users interested in different events in diverse contexts. [ABSTRACT FROM AUTHOR]