Forecasting collective behavior using social media
Authors
Korolov, Rostyslav
ORCID
Loading...
Other Contributors
Wallace, William A., 1935-
Mendonça, David
Ji, Heng
Pequito, Sérgio
Mendonça, David
Ji, Heng
Pequito, Sérgio
Issue Date
2018-12
Keywords
Decision sciences and engineering systems
Degree
PhD
Terms of Use
Attribution-NonCommercial-NoDerivs 3.0 United States
This electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
This electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
Full Citation
Abstract
This thesis constructs hypotheses and measurements based on theoretical explanations of collective behaviors using a modeling framework informed by Emergent Norm Theory. Developed long before the Internet, the theory explains collective behaviors as a process of adoption of new social norms that emerge collectively in response to unanticipated or ambiguous events. New norms are propagated through norming acts, which take the form of verbal communication and non-verbal actions. Emergent norms also manifest in changes to extant communication patterns and roles in a collective. This study examines the applicability of Emergent Norm Theory to social media communications, with the primary hypothesis being that social media messaging in response to unexpected events is a norming act and thus indicates emergence of new norms that lead to collective behaviors. Another hypothesis is that the evolution of a social network observed in social media is a manifestation of a change to communication patterns and roles associated with emergent norms. Three empirical studies test these hypotheses using disparate aspects of online communications as measures of constructs associated with Emergent Norm Theory. The measures are then compared to observed behaviors.
Online social media gained worldwide popularity during the first decade of this century, attracting attention from both industry and academia, with vast amounts of recorded written communication the primary reason for the interest. One pertinent use of such data is studying large-scale social processes, reflected in human communications over the Internet. This thesis uses online social media data to hindcast social processes inherent in collective behaviors in response to external events. Some research forecasts using social media data, with subjects of such forecasts including flu epidemics and election results. With few exceptions, such research uses observed patterns in data without addressing theoretical explanations.
A study that predicts the value of charitable donations uses aggregate measures of (1) the amount of communication and (2) social network structures to make predictions. Assessing social unrest, the second study uses temporal dynamics of the amount of communication to predict observed instances of protest. Also predicting social unrest, the third study uses temporal dynamics of social network structures, disregarding the content of communication, to predict occurrences of protest. This thesis demonstrates three approaches to predicting collective behaviors using social media data, informed by theoretical studies of collective behavior. Results suggest that Emergent Norm Theory applies to studying collective behaviors using social media as data sources, providing insights into the role of social networks in propagation of collective behaviors. It also suggests implications for theories that explain collective behaviors in the context of online social media. Results and the methods used to obtain them are of practical use in topics such as emergency response management, public safety, and other public services.
Online social media gained worldwide popularity during the first decade of this century, attracting attention from both industry and academia, with vast amounts of recorded written communication the primary reason for the interest. One pertinent use of such data is studying large-scale social processes, reflected in human communications over the Internet. This thesis uses online social media data to hindcast social processes inherent in collective behaviors in response to external events. Some research forecasts using social media data, with subjects of such forecasts including flu epidemics and election results. With few exceptions, such research uses observed patterns in data without addressing theoretical explanations.
A study that predicts the value of charitable donations uses aggregate measures of (1) the amount of communication and (2) social network structures to make predictions. Assessing social unrest, the second study uses temporal dynamics of the amount of communication to predict observed instances of protest. Also predicting social unrest, the third study uses temporal dynamics of social network structures, disregarding the content of communication, to predict occurrences of protest. This thesis demonstrates three approaches to predicting collective behaviors using social media data, informed by theoretical studies of collective behavior. Results suggest that Emergent Norm Theory applies to studying collective behaviors using social media as data sources, providing insights into the role of social networks in propagation of collective behaviors. It also suggests implications for theories that explain collective behaviors in the context of online social media. Results and the methods used to obtain them are of practical use in topics such as emergency response management, public safety, and other public services.
Description
December 2018
School of Engineering
School of Engineering
Department
Dept. of Industrial and Systems Engineering
Publisher
Rensselaer Polytechnic Institute, Troy, NY
Relationships
Rensselaer Theses and Dissertations Online Collection
Access
CC BY-NC-ND. Users may download and share copies with attribution in accordance with a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License. No commercial use or derivatives are permitted without the explicit approval of the author.