Representing and analyzing location-based social media activity in GIS
MetadataShow full item record
Human activities are generated due to physiological, psychological and economic needs, and have spatial, temporal, and social elements. In the era of big data, geo-tagged social media are becoming new platforms that influence human behaviors in space and time, and are also serving as new channels for geographers to observe human interactions and social connections at fine scales. Traditional geographical representation and analysis methods in GIS are not sufficient to tackle the much more complex nature of the location-based social media activity. The convergence of GIS and social media has resulted in data avalanche and requires new theories in GIScience. This dissertation has developed methods and tools to represent and analyze location-based social media activates (LBSMA) in GIS in three perspectives: (1) this research has proposed a conceptual framework of location-based social media activity to model human activities in spatial-temporal-social dimension , and has implemented this data framework to organize LBSMA in GIS and produce practical tools to calculate useful measurements; (2) this research has developed a random walking algorithm to characterize urban road networks by calculating the possibility distribution of human locations over time; and (3) this research has introduced location-based social connections to visualize and quantify social connections in spatial-temporal dimension. In addition, this research has established a dedicated website (www.lbsoical.net) to extract and analyze the real-world social media data, and provide visualization and analysis function for various studies. The developed methods and tools in this research can organize, visualize, simulate and analyze human activities in spatial-social-temporal dimension. Those methods have added to our understanding of human interactions by providing innovative and applicable measures for places, social connections and human activities. The findings from this research have yielded new insights regarding human activities in virtual and physical space, and have enhanced technical capabilities for social media analysis in GIS. The developed methods can help identify place-based or people-based strategies, e.g., urban planning, traffic planning, commercial advertising or energy communicating. The proposed framework will pave new avenues for future research, such as public health, transportation, urban geography and social science.