Online social networking sites carry a lot of information organized around the people who actually submit it. They rely a lot on the interconnections between the actors of the network and they relate the generated information in terms of connections or ties among those actors. On the other hand, social networks carry a lot of information in real time or almost real time since people "report" information as they see it unfolding before their eyes. We are targeting this information and we try to understand when this information can be related to refer to a single event, which has both a spatial and a temporal dimension. These social sensors can alert us on what is happening in the society when it happens and by aggregating the different reports from people who are on the scene. In order to achieve that we suggest and describe a set of services that can be used for collection of the information, identifying the discussion topics and provide at the end an alert if the discussion qualifies for and contains enough information to define an event. The goal is for the system to work with no prior knowledge of the events we seek.