摘要
众包产生于比较复杂的互联网平台上,必须对互联网平台上的众包质量进行控制,研究基于社交平台的众包质量控制算法尤为必要。根据众包问题涉及领域,将用户在社交平台领域的直接信誉度算法与用户对历史任务完成情况的质量评估算法相结合完成用户筛选,并根据筛选用户给出的方案集,利用最大期望算法(E-M算法)获取正确率相对较高的方案。实验结果表明,即使在加入了一些恶意工作者的情况下,利用直接信誉度算法与用户质量评估算法筛选用户,并使用E-M算法处理方案集能够使社交平台上的众包质量得到较好控制。
As crowd-sourcing is generated on the Internet platform complex relatively,it is necessary to control the quality of crowd-sourcing on the Internet platform.So far,however,there has been little research into crowd-sourcing quality control on social platforms.Mainly studies the quality control algorithm of crowd-sourcing based on social platform.Firstly,this paper adopts the user＇s direct reputation algorithm based on the social platform and the user＇s quality evaluation algorithm for the completion of the historical task to filter users,according to the domain covered of crowd-sourcing problem.Finally,according to the scheme set of the filtered users,the maximum expectation algorithm（EM algorithm）is adopted to obtain the scheme with correct rate relatively high.The experimental results show that,even in the case of some malicious workers joining in,using the direct algorithm of the reputation and the quality of the user evaluation algorithm to filter users,and using EM algorithm to process scheme set can make the quality of crowd-sourcing on social platform get control better.
出处
《软件导刊》
2017年第12期90-93,共4页
Software Guide
关键词
众包
社交平台
质量检测
领域信誉度
最大期望算法
crowdsourcing
social networking platform
quality inspection
field of credibility
maximumexpected algorithm
