Visible to the public S&CC: Support Community-Scale Study by Visual Analytics of Human Mobility and Opinion Data from Social Media DataConflict Detection Enabled

Project Details
Lead PI:Ye Zhao
Co-PI(s):Jing Yang
Xinyue Ye
Andrew Curtis
Performance Period:08/15/16 - 07/31/18
Institution(s):Kent State University
Sponsor(s):National Science Foundation
Award Number:1637242
731 Reads. Placed 497 out of 804 NSF CPS Projects based on total reads on all related artifacts.
Abstract: Improvements to the urban physical landscape, such as adding new greenspaces and healthy activity destinations, or removing problem areas like "blight" are decided upon using ideas, plans and existing theory. Yet little is ever done to evaluate the effectiveness of such improvements. The project develops a visual analytics platform of anonymized human mobility and human opinion data retrieved from social media, so that community-focused stakeholders can interactively study human activity usage and associate insights around multiple location types. In this way questions can be asked such as has this change actually done any good on this/my street? Or, where can I tell my patient to exercise that is safe, culturally acceptable, and appropriate to who he/she is? This visual analytics system can become part of a best practices approach when re-designing urban environments, where re-designing means changing aspects of the built, social and physical environment. The Broader Impacts of this work include potential usage beyond academia; a doctor interactively investigating the neighborhood around a patient, or a community group considering the likely impact of adding a community garden by investigating the changes such an initiative has had elsewhere. The project provides a non-specialist software interface that makes that type of analysis ubiquitous. This tool can be applied irrespective of cultural, racial or economic barriers. The geographic area of the proposed study comprises largely of minority neighborhoods allowing us to show the utility of this approach to attack issues of racial disparity. Every effort will be made to attract and employ minority students on the project in order to reflect the social makeup of these study neighborhoods. This project is the first near real time intervention tool designed for community use that utilizes the most dynamic data available: social media trajectory data and associated contextual meaning. The technique takes social media and leverages it in an interactive visual system to answer day-to-day questions by non-specialized users. It captures behaviors/activities related to urban communities by investigating and understanding social media trajectory data. These points of investigation will be gleaned from a rich array of sources including community narratives. Important negative and positive spaces, and changes to those spaces, will be investigated to justify the utilization of big social media data and evaluate our visual analytics system. A database, NeighborBase, manages these heterogeneous data extracted from tweets and support various real-time queries. A community-scale visual analytics tool, NeighborVis, further provides intuitive interactions for users to perform efficient knowledge discovery over NeighborBase. The software platform seamless integrates these data with computational and visualization algorithms for the design and improvement in physical communities.