Smart government framework with geo-crowdsourcing and social media analysis

Abstract Smart government utilizes data to improve services provided to the citizen. Smart city may benefit from citizen engagement through crowdsourcing and social media. This paper proposes a city event’s prediction with map-based visualization utilizing crowdsourced geodata collected from citizen’s report. Data is aggregated according to geo-position and used to predict next day’s event. Mindmap-based visualization utilizing social media processed with topic modeling and topic similarity calculation is also proposed in this work. Proof-of-concept application and proposed system’s processing time is evaluated to prove the feasibility of the system. Proposed system utilizes geo-crowdsourcing for event prediction visualization for government instead of direct services to citizen. Proposed social media visualization use topic modeling specifically created for short text which is appropriate for social media data.