Approximation Algorithms for Scheduling Crowdsourcing Tasks in Mobile Social Networks

TL;DR

This paper corrects previous analysis errors and introduces new randomized and deterministic approximation algorithms for scheduling tasks in mobile social networks, achieving improved approximation ratios for minimizing total weighted completion time.

Contribution

It provides the correct approximation ratio analysis for existing algorithms and proposes new algorithms with better performance guarantees for task scheduling.

Findings

Corrected the approximation ratio analysis for existing scheduling algorithms.

Introduced a randomized algorithm with an expected ratio of 1.5 + ε.

Presented a deterministic algorithm with a ratio of max{2.5, 1+ε}.

Abstract

This paper addresses the scheduling problem in mobile social networks. We begin by proving that the approximation ratio analysis presented in the paper by Zhang \textit{et al.} (IEEE Transactions on Mobile Computing, 2025) is incorrect, and we provide the correct analysis results. Furthermore, when the required service time for a task exceeds the total contact time between the requester and the crowd worker, we demonstrate that the approximation ratio of the Largest-Ratio-First task scheduling algorithm can reach $2 - \frac{1}{m}$. Next, we introduce a randomized approximation algorithm to minimize mobile social networks' total weighted completion time. This algorithm achieves an expected approximation ratio of $1.5 + \epsilon$ for $\epsilon>0$. Finally, we present a deterministic approximation algorithm that minimizes mobile social networks' total weighted completion time. This deterministic algorithm achieves an approximation ratio of $\max\left\{2.5,1+\epsilon\right\}$ for $\epsilon>0$. Additionally, when the task's required service time or the total contact time between the requester and the crowd worker is sufficiently large, this algorithm can reach an approximation ratio of $1.5+\epsilon$ for $\epsilon>0$.