摘要
越来越多的公司或企业借助众包来完成工作,导致众包任务数量急剧增长,而当前众包平台的任务搜索功能还不能满足工人的个性化需求,这使工人难以从海量任务中选择感兴趣且擅长的任务。针对众包工作模式的特点,提出了一种基于工人—任务关系评价兴趣度值的方法,并结合工人的浏览内容和历史完成记录提出了用户偏好模型的表示和更新机制,从而建立用户兴趣模型。选取在众包平台上的真实数据集进行实验分析,验证了该模型的性能,证明该模型能较为准确的反映工人的兴趣偏好。
More companies or enterprises complete their work by crowdsourcing, resultcrease of crowdsourcing tasks. However, the current crowdsourcing platform cannot meet for personalized task searching, which makes workers feel more difficutt to sort out the suitable tasks from the massivetasks. This paper proposed a method to evaluate the interest rate based on the relatioand task for crowdsourcing. Combining the contents the user had browsed with the history records, wepresent the representation and update mechanism for user interest model. The effectiveness of the model is verified through experimental analysis on real dataset of crowdsourcing platform. The resutt shows thatthis model can reflect the interest of worker with a high rate of accuracy.
作者
王晓燕
李劲华
WANG Xiao- yan, LI Jin- hua(School of Data Science and Software Engineering, Qingdao University, Qingdao 266071, Chin)
出处
《青岛大学学报（自然科学版）》
CAS
2018年第1期102-108,共7页
Journal of Qingdao University(Natural Science Edition)
关键词
众包
任务搜索
个性化
兴趣度
用户偏好模型
crowdsourcing
task searching
personalized
interest ra te
user interest model
