| 引用本文: | 张琳,姜高霞,王文剑.众包标签的双重置信度推断算法.[J].国防科技大学学报,2022,44(3):77-84.[点击复制] |
| ZHANG Lin,JIANG Gaoxia,WANG Wenjian.Crowdsourced label inference algorithm using double-confidence[J].Journal of National University of Defense Technology,2022,44(3):77-84[点击复制] |
|
| |
|
|
| 本文已被：浏览 4429次 下载 3613次 |
| 众包标签的双重置信度推断算法 |
| 张琳,姜高霞,王文剑 |
|
(山西大学 计算机与信息技术学院, 山西 太原 030006)
|
| 摘要: |
| 标记者的知识水平、评价标准等均具有显著差异,导致收集到的标签质量参差不齐,提高标签和学习模型的质量对众包标签中学习起着关键作用。针对众包标签推断问题,提出了一种双重置信度推断算法,分别从数据分布特征及标签信息两方面计算得到标记者置信度,再通过此置信度推断数据集的集成标签,以此提高集成标签的质量。实验结果表明,与其他仅使用标签信息的推断算法相比,所提算法可以得到更优结果。 |
| 关键词: 众包 标签集成 聚类 双重置信度 |
| DOI：10.11887/j.cn.202203010 |
| 投稿日期：2021-08-30 |
| 基金项目:国家自然科学基金资助项目(62076154,61906113,U1805263)；山西省国际合作重点研发计划资助项目(201903D421050)；山西省高等学校科技创新资助项目(2020L0007)；中央引导地方科技创新资助项目(YDZX20201400001224) |
|
| Crowdsourced label inference algorithm using double-confidence |
| ZHANG Lin, JIANG Gaoxia, WANG Wenjian |
|
(School of Computer and Information Technology, Shanxi University, Taiyuan 030006, China)
|
| Abstract: |
| Since the workers have significant differences in the knowledge level and evaluation criteria, the quality of the collected labels varies a lot. It′s of key importance to improve the quality of labels and learning models in crowdsourced label learning. A novel double-confidence inference algorithm was proposed to solve the problem of crowdsourced label inference. The workers′ confidence was obtained via the data distribution characteristics and label information, and then the label was inferred by this confidence so as to improve the quality of the integrated label. The experimental results show that the proposed algorithm outperforms other ground truth inference algorithms only based on label information. |
| Keywords: crowdsourced integrated label clustering double-confidence |
|
|
|
|
|
