Spatial crowdsourcing based on Web mapping services

Abstract

With the rapid development of mobile Internet and smartphones, crowdsourcing marketplaces has extended to spatial crowdsourcing (SC), where crowd workers perform spatial tasks (i.e., tasks related to a location) in the physical world using their mobile phones. Currently, most of existing spatial crowdsourcing algorithms (e.g., task assignment) assume the underlying road network is given or simply base on Euclidean space. However, in fact, not every spatial crowdsourcing platform has enough resources to possess map data (e.g., the road network and live traffic information) by itself, especially for these that are small or startup companies, while the Euclidean distance is usually not accurate enough for SC processing. To overcome these limitations, we propose a spatial crowdsourcing system based on Web mapping services, i.e., the spatial crowdsourcing platform can subscribe distance, live travel time and detailed route information from Web mapping services through their APIs, and utilize these retrieved map data for SC processing; furthermore, workers and task requesters can also snap their real locations to the locations on the road for privacy protection through the APIs. As retrieving map data from Web mapping services is much more expensive than accessing local data, we take the advantage of the pruning and route sharing approaches to reduce the number of external requests to Web mapping services, and our experimental results have proved their effectiveness.