Optimizing Microtask Assignments on Crowdsourcing Platforms Using Jaro-Winkler
Abstrak
The increased utilization of crowdsourcing platforms in distributing microtasks has become a significant trend in information systems. However, the main challenge lies in optimizing the efficiency of microtasking to enhance the quality of outcomes while minimizing time and costs. In this research, we explore the Jaro-Winkler method to enhance the performance of microtasking on crowdsourcing platforms. The Jaro-Winkler method is employed to measure the similarity between tasks and workers' capabilities, enabling more accurate filtering in microtask assignments. We implement this model in the SME sector, specifically in Talent Services, and observe a significant improvement in task completion speed and result accuracy. The research stages encompass preliminary research, conceptual model design, data collection, method implementation, job quality estimation, and model validation. The findings make a vital contribution to the literature on microtask optimization and provide a fresh perspective on integrating string matching technology in the crowdsourcing context. This research successfully aids SMEs in reducing assignment errors, allowing for more precise matching between task descriptions and worker capabilities
Unduhan
Diterbitkan
Cara Mengutip
Terbitan
Bagian
Lisensi
Hak Cipta (c) 2023 Proceeding KONIK (Konferensi Nasional Ilmu Komputer)
Artikel ini berlisensi Creative Commons Attribution-NonCommercial 4.0 International License.