The Importance of Worker Reputation Information in Microtask-Based Crowd Work Systems

TL;DR

This paper investigates how incorporating worker reputation information can significantly enhance the efficiency and accuracy of microtask crowd work systems, proposing algorithms and analyzing their effectiveness.

Contribution

It formalizes the optimal task assignment problem with reputation data, proposes heuristic algorithms, and evaluates their performance in crowd work systems.

Findings

Reputation information improves system performance even when estimates are inaccurate.

Maximum a-posteriori decision rule performance declines with reputation estimate inaccuracy.

Combining task assignment with message-passing decision algorithms yields best results under high inaccuracy.

Abstract

This paper presents the first systematic investigation of the potential performance gains for crowd work systems, deriving from available information at the requester about individual worker reputation. In particular, we first formalize the optimal task assignment problem when workers' reputation estimates are available, as the maximization of a monotone (submodular) function subject to Matroid constraints. Then, being the optimal problem NP-hard, we propose a simple but efficient greedy heuristic task allocation algorithm. We also propose a simple "maximum a-posteriori" decision rule and a decision algorithm based on message passing. Finally, we test and compare different solutions, showing that system performance can greatly benefit from information about workers' reputation. Our main findings are that: i) even largely inaccurate estimates of workers' reputation can be effectively exploited in the task assignment to greatly improve system performance; ii) the performance of the maximum a-posteriori decision rule quickly degrades as worker reputation estimates become inaccurate; iii) when workers' reputation estimates are significantly inaccurate, the best performance can be obtained by combining our proposed task assignment algorithm with the message-passing decision algorithm.