Human dynamics in the age of big data: a theory-data-driven approach

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Nastaran Pourebrahim (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site:
Selima Sultana

Abstract: The revolution of information and communication technology (ICT) in the past two decades have transformed the world and people’s lives with the ways that knowledge is produced. With the advancements in location-aware technologies, a large volume of data so-called “big data” is now available through various sources to explore the world. This dissertation examines the potential use of such data in understanding human dynamics by focusing on both theory- and data-driven approaches. Specifically, human dynamics represented by communication and activities is linked to geographic concepts of space and place through social media data to set a research platform for effective use of social media as an information system. Three case studies covering these conceptual linkages are presented to (1) identify communication patterns on social media; (2) identify spatial patterns of activities in urban areas and detect events; and (3) explore urban mobility patterns. The first case study examines the use of and communication dynamics on Twitter during Hurricane Sandy utilizing survey and data analytics techniques. Twitter was identified as a valuable source of disaster-related information. Additionally, the results shed lights on the most significant information that can be derived from Twitter during disasters and the need for establishing bi-directional communications during such events to achieve an effective communication. The second case study examines the potential of Twitter in identifying activities and events and exploring movements during Hurricane Sandy utilizing both time-geographic information and qualitative social media text data. The study provides insights for enhancing situational awareness during natural disasters. The third case study examines the potential of Twitter in modeling commuting trip distribution in New York City. By integrating both traditional and social media data and utilizing machine learning techniques, the study identified Twitter as a valuable source for transportation modeling. Despite the limitations of social media such as the accuracy issue, there is tremendous opportunity for geographers to enrich their understanding of human dynamics in the world. However, we will need new research frameworks, which integrate geographic concepts with information systems theories to theorize the process. Furthermore, integrating various data sources is the key to future research and will need new computational approaches. Addressing these computational challenges, therefore, will be a crucial step to extend the frontier of big data knowledge from a geographic perspective. KEYWORDS: Big data, social media, Twitter, human dynamics, VGI, natural disasters, Hurricane Sandy, transportation modeling, machine learning, situational awareness, NYC, GIS

Additional Information

Language: English
Date: 2019
Big data, GIS, Natural disasters, Social media, Transportation modeling, VGI
Big data
Social media
Social interaction
Spatial behavior
Information technology

Email this document to