Fusing social network data with hard data

Social networking sites such as Twitter, Facebook and Flickr play an important role in disseminating breaking news about natural disasters, terrorist attacks and other events. They serve as sources of first-hand information to deliver instantaneous news to the masses since millions of users visit these sites to post and read news items regularly. Hence, by exploring efficient mathematical techniques like Dempster- Shafer theory and Modified Dempster's rule of combination, we can process large amounts of data from these sites to extract useful information in a timely manner. In surveillance related applications, the objective of processing voluminous social network data is to predict events like revolutions and terrorist attacks before they unfold. By fusing the soft and often unreliable data from these sites with hard and more reliable data from sensors like radar and the Automatic Identification System (AIS), we can improve our event prediction capability. In this paper, we present a class of algorithms to fuse hard sensor data with soft social network data (tweets) in an effective manner. Preliminary results using real AIS data are also presented.

I. Introduction

In defence, military or homeland security systems in order to track and predict events, and to track mobile target states, the decision makers need accurate data. Due to limited fields-of-view and obscuration, conventional prediction and tracking methods [2], [3] that rely exclusively on hard sensors (e.g., radar, sonar, video) can make erroneous decisions. On the other hand, algorithms that use only soft data (e.g., human input, social network data) can be ineffective due to conflicting unreliable information. In some cases, the unreliability of soft data might be intentional. Social Network (SN) data is one form of soft data that has many advantages: it is voluntary, voluminous, instantaneous and evolving. As a result, it is a rich source of data that is contributed over time by a large number of identifiable users, who are often close to unfolding events of interest, at virtually no cost to us. This has spurred great interest in mining social data for information extraction and exploitation. Specifically, the fusion of soft and hard data is of significant interest in many surveillance systems. This indeed provides the motivation for the proposed work.