With the rapid development of sharing economy and mobile Internet in recent years, a wide range of applications of the Two-sidedOnlineBipartiteMatching (TOBM) problem in spatial data are gaining increasing popularity. To be specific, given a group of workers and tasks that dynamically appear in a 2D space, the TOBM problem aims to find a matching with the maximum cardinality between workers and tasks satisfying the spatiotemporal constraints. Many works have studied this problem, but the settings of their problems are different from each other. Moreover, no prior works have compared the performances of the algorithms tailored for different settings under a unified definition. As a result, there lacks a guideline for practitioners to adopt appropriate algorithms for various scenarios. To fill the blank in this field, we present a comprehensive evaluation and analysis of the representative algorithms for the TOBM problem in this paper. We first give our unified definition and then provide uniform implementations for all the algorithms. Finally, based on the experimental results on both synthetic and real datasets, we discuss the strengths and weaknesses of the algorithms in terms of short-term effect and long-term effect, which can be guidance on selecting appropriate solutions or designing new methods.

中文翻译：

---

空间数据中的双面在线二分匹配：实验和分析

随着经济的共享和移动互联网在近几年的快速发展，各种各样的应用牛逼WO片面Ø n线乙ipartite中号atching（TOBM）空间数据中的问题越来越受欢迎。具体来说，给定一组在2D空间中动态出现的工作人员和任务，TOBM问题旨在找到满足时空约束的工作人员和任务之间具有最大基数的匹配。许多作品都研究了这个问题，但是它们的问题背景各不相同。而且，在统一的定义下，没有任何先前的工作可以比较针对不同设置量身定制的算法的性能。结果，缺乏针对从业者针对各种场景采用适当算法的指南。为了填补这一领域的空白，本文对TOBM问题的代表性算法进行了全面的评估和分析。我们首先给出统一的定义，然后为所有算法提供统一的实现。最后，基于合成和真实数据集的实验结果，我们从短期效果和长期效果的角度讨论了算法的优缺点，可以为选择合适的解决方案或设计新方法提供指导。