Seeking an Online Social Media Radar

Seeking an Online Social Media Radar

James ter Veen (The George Washington University, USA), Shahram Sarkani (The George Washington University, USA) and Thomas A. Mazzuchi (The George Washington University, USA)
Copyright: © 2015 |Pages: 26
DOI: 10.4018/978-1-4666-8556-7.ch005


In this paper we identify a method, which rapidly analyzes vast amounts of data present in social media in order to forecast crowd sizes. Based upon comparative analysis of related literature, a conceptual model is proposed and research conducted to develop capabilities to forecast mass collective action behavior such as crowd formation using Social Network Analysis (SNA) tools applied to online social media. We demonstrate that a simple model of online social network parameters can produce situation awareness of crowd sizes in much the same way that radar sensors can produce situation awareness of air traffic density. A prototype online social media ‘radar' sensor system is developed and tested in a pilot study with a dataset of tweets gathered regarding the Occupy Wall Street movement. Further work is suggested which could provide anticipated crowd location, movement and intent in addition to size.
Chapter Preview

Background: Sensor System Requirements

If the sensor to be evaluated were a new radar system, requirements might include a set of statements such as 'the radar beam shall sweep through 360 degrees of airspace. One complete rotation of the beam shall be accomplished in no more than 5 seconds. Positional error of radar plots shall be no greater than 20%.' For an online social media ‘radar’ system, we derive a high-level set of generic system requirements shown in Table 1 (for the purposes of prototype development, this is not an exhaustive set of all possible system requirements) from published very high-level electronic radar system specifications (Lockheed Martin, 2010; Thales Raytheon Systems, 2011; Wolff, 2009). First we need to define ‘online social media radar’. Maybury (2010) defines a ‘social radar’ as a tool that “needs to sense perceptions, attitudes, beliefs and behaviors” and that it would use signatures “and correlation to sense, if not forecast, a broad spectrum of phenomena” including changing trends in population behaviors. For the purposes of this study, we initially define our prototype ‘online social media radar’ as a software tool, which uses correlation of data gathered from online social media to sense and forecast sizes of crowd formations.

Complete Chapter List

Search this Book: