Hyperspatial crowdsourcing is a low-cost methodology that uses capable individuals, located in the vicinity of task location, to acquire fine-grain data for spatially distributed tasks. However, it is constrained by the budget. This paper analyzes different scenarios presented by the arrival and departure of individuals from the task arena. It introduces a taxonomy of several problem variants. It studies the hardness of task assignment problems for offline analysis, and presents online heuristics that use acquired spatial and temporal knowledge. These are tested through a set of experiments.

The paper first introduces the topic of spatial crowdsourcing and introduces the concept of hyperspatial crowdsourcing. Then it proceeds to a survey of related work and formally defines related concepts such as task, worker, worker-task coverage, and problem taxonomy for budget-per-time-period versus budget-per-campaign, binary utility versus distance-based utility, single-objective versus multi-objective, and offline versus online. In subsequent sections, the hardness of the problem is defined and analyzed for fixed budget (fMTC) and dynamic budget (dMTC). A basic algorithm for fMTC and another for dMTC are presented.

A short discussion on issues of worker overload and performance evaluation for algorithms is followed by experimental results. The authors note that they do not consider the mobility of a group of workers; however, they conclude that the proposed algorithms and procedure are an improvement over existing spatial crowdsourcing methods.

The paper has 48 references and would interest readers in the spatial crowdsourcing field.