不同风险偏好下两类众包机制最优决策

发布时间：2018-05-01 11:39

本文选题：众包 + 完全支付拍卖　；参考：《系统工程》2017年05期

【摘要】：研究固定悬赏金额的众包任务中,在发布者最大化任务总质量和最高质量两类目标机制下,解答者风险偏好对发布者获得质量和解答者期望效用的影响。首先,用完全支付拍卖模型求解不同风险偏好下,解答者期望效用和发布者获得质量。其次,运用Stackelberg博弈的方法对两种目标机制的最优决策进行求解及比较分析,发现在两类目标机制中均有:解答者期望效用随解答者人数(风险偏好)的增加而减少(增加),发布者获得的解答质量随解答者人数(风险偏好)的增加而增加;但当解答者数量无穷大时,两者均趋于定值。然后,使用Taskcn.com网站数据对模型的推论进行检验,除推论5、推论6被部分支持外,其他推论均获得显著支持。最后,拓展研究表明随着参与经验的积累,解答者的风险偏好显著变小。
[Abstract]:In a crowdsourcing task with a fixed reward amount, under the objective mechanism of maximizing the total quality of the task and the highest quality, the effect of the risk preference of the solution on the quality of the publisher and the expected utility of the solution is studied. Firstly, the expected utility of the solution and the quality of the publisher are obtained by using the full payment auction model to solve different risk preferences. Secondly, the Stackelberg game method is used to solve and compare the optimal decision of the two kinds of target mechanism. It is found that there are two kinds of target mechanisms: the expected utility of the respondent decreases with the increase of the number of respondents (risk preference) and the quality of the solution obtained by the publisher increases with the increase of the number of respondents (risk preference); But when the number of answers is infinite, both tend to be fixed. Then, using Taskcn.com website data to test the inference of the model, except inference 5, inference 6 is partially supported, all other inferences are supported significantly. Finally, the extended study shows that the risk preference of the responders decreases significantly with the accumulation of experience.
【作者单位】：南开大学商学院;青岛农业大学经济学院;山东财经大学管理科学与工程学院;太原理工大学经济管理学院;
【基金】：国家社会科学基金资助项目(15BGL074);国家社会科学基金青年项目(17CGL019) 山东省自然科学基金博士项目(ZR2017BG010) 山东省高等学校人文社科计划项目(J17RB094)
【分类号】：F224.32

【相似文献】