Exploiting Heterogeneous Data Sources: A Computing Paradigm for Live Web and Sustainability Applications

Abstract

Today we are witnessing advances in technology that are changing dramatically the way we live, work and interact with the physical environment. New revolutionary technologies are creating an explosion on the size and variety of information that is becoming available. Such technologies include the development and widespread adoption of networks of small and inexpensive embedded sensors that are being used to instrument the environment at an unprecedented scale. In addition, the last few years have brought forward the widespread adoption of social networking applications. Another trend with significant ramifications is the massive adoption of smartphones in the market. The rise of the social networking applications and the always-on functionality of the smartphones are driving the rise of a part of the web that is dedicated to recording, maintaining and sharing rapidly changing data which has been termed the Live Web. In this talk we present recent research work motivated by the trends we describe above. We also consider how such novel research results are enabling forms of computation. First, we focus on the specific problem of finding events or trends, including spatiotemporal patterns, when monitoring microblogging streams. Our work is mainly in the context of the INSIGHT FP7 project and we also consider data from sources as different as traffic sensors and Twitter streams. To put this research work in a general context, in the second part of the talk we consider the more general problem of developing applications and reasoning about the behavior of novel applications that exploit the new setting of the Live Web, and understanding the implications on the design, development and deployment of new applications in this setting. We describe initial work on the formulation of a new computing paradigm for this setting, and on describing how it can be applied for computational sustainability applications.