The design of puzzle selection strategies for ESP-like GWAP systems

Ling-Jyh Chen*, Bo Chun Wang, Wen Yuan Zhu

*此作品的通信作者

研究成果: 雜誌貢獻期刊論文同行評審

5 引文 斯高帕斯(Scopus)

摘要

The games with a purpose (GWAP) genre is a type of human computation that outsources certain steps of the computational process to humans. Although most GWAP studies focus on the design and analysis of GWAP systems, a systematic and thorough evaluation of existing systems is lacking. We address the issue in this paper. Taking the ESP game as an example, we propose a metric, called system utility, for evaluating the performance of GWAP systems, and use analysis to study the properties of the ESP game. We argue that GWAP systems should be designed and played with strategies. To this end, based on our analysis, we implement an optimal puzzle selection strategy (OPSA) to improve GWAP systems. Using a comprehensive set of simulations, we show that the proposed OPSA approach can improve the system utility of the ESP game significantly. In addition, we implement a quasi ESP game, called ESP Lite, which embeds three puzzle selection algorithms transparently and records the complete game trace for evaluation and further research. During a one-month experiment, we have investigated the inner properties of the three strategies in real-world GWAP systems, and verified that the OPSA scheme achieves the best system utility for the ESP game. The results of this study demonstrate that GWAP systems are more efficient if they are designed and played with strategies.

原文英語
文章編號5460962
頁(從 - 到)120-130
頁數11
期刊IEEE Transactions on Computational Intelligence and AI in Games
2
發行號2
DOIs
出版狀態已發佈 - 2010 六月

ASJC Scopus subject areas

  • 軟體
  • 控制與系統工程
  • 人工智慧
  • 電氣與電子工程

指紋

深入研究「The design of puzzle selection strategies for ESP-like GWAP systems」主題。共同形成了獨特的指紋。

引用此