Integrating big social data, computing and modeling for spatial social science

In the past decade, the social sciences are undergoing a dramatic shift toward analyzing ever-increasing amounts of large-scale diverse data with rich spatial, temporal, and thematic resolution (Shaw, Tsou, and Ye 2016). With the introduction of new data, methods and computing platforms, rigorous analysis of emerging socioeconomic events opens up a rich empirical context for the social sciences and policy interventions. Due to the rapid progress of data acquisition tools and Internet techniques, a new period called the “big data era” has appeared and exerted significant influence on social science research. This special section aims to shed light on the opportunities, challenges, and solutions of integrating the big social data and latest computing, modeling and information techniques for spatial social sciences. This special section contains five articles and rejected four manuscripts; each has been selected through a rigorous peer-review process. The first paper, “SinoGrids: a practice for open urban data in China” (Zhou and Long), initiates a crowdsourcing platform that shares multisource microscale urban data to facilitate efficient and effective urban analytics. Through downscaling the data based on a discrete uniform grid, this platform (SinoGrids) addresses the following three major issues: the balance between benefits to original data holders and data users, the gap from accessible data to effectively usable data, and the failure to share data by many data users. Focusing on data challenges for urban studies, this research can promote the open data movement in China to make more publicly acquired data available at the local, regional, and national scales. The second paper, “Crowdsourcing functions of the living city from Twitter and Foursquare data” (Zhou and Zhang), extracts and analyzes six types of human activities (education, dining, nightlife, shopping, travel & transportation, and recreation) from Twitter and Foursquare data to identify urban functions in a realtime system. This research illustrates that crowdsourced social media data can be utilized to detect when and where these human activities occur. With the growing availability of large road network data such as OpenStreetMap and TIGER/Line, selective omission is essential for fully automated transformation of a map from one scale to a smaller scale. The third paper, “How many samples are needed? An investigation of binary logistic regression for selective omission in a road network” (Zhou and Li), employs the supervised learning approach in road selection and investigates the sample section issue. Various percentages and numbers of strokes are used as samples for training a supervised learning approach, before being applied to the untrained strokes for validation. The fourth paper, “Travel time estimation at intersections based on low-frequency spatial-temporal GPS trajectory big data” (Tang et al.), estimates the average intersection travel time of different vehicles through an intersection, based on Global Positioning System (GPS) vehicle trace data from taxicabs using a fuzzy fitting approach. The proposed method is appropriate for estimating intersection travel time in a short period of time, based on road network data and GPS trace big data from 11,283 taxicabs. The fifth paper, “The socio-environmental data explorer (SEDE): a social media–enhanced decision support system to explore risk perception to hazard events” (Shook and Turner), develops a prototype social media–enhanced decision support system as a synergistic integration of social media and decision support systems, in order to evaluate social response to environmental risk and hazard events. This system has been utilized to enhance situational awareness in a case study mapping response to an extreme weather event in the northeast of the United States. Even though the maturing Web technology has made social data largely available on the Internet, it is still very challenging to discover and access large volumes of data due to variety in formats and complexity in encoding. Investigating complex human and socioeconomic dynamics needs a platform upon which effective solutions can be possibly developed in an interdisciplinary, collaborative, and timely manner. Users of social media services frequently update their status and post tweets (and/or pictures) online, some of which could be shared exponentially. These real-time data, complemented with official and authoritative data sources, are valuable, yet require a computing environment that is adaptable to expand and store considerable amounts of data in timely manner. CARTOGRAPHY AND GEOGRAPHIC INFORMATION SCIENCE, 2016 VOL. 43, NO. 5, 377–378 http://dx.doi.org/10.1080/15230406.2016.1212302