Homophily-aware correction framework for crowdsourced labels using heterogeneous information network

作者：

Highlights：

• A heterogeneous information network-based method is proposed for label correction.

• Homophily among labelers is defined to improve the quality of crowdsourced labels.

• A homophily-based classifier is proposed to enhance impacts of positive labels.

• Meta-paths are employed to capture implicit relations between labelers.

摘要

•A heterogeneous information network-based method is proposed for label correction.•Homophily among labelers is defined to improve the quality of crowdsourced labels.•A homophily-based classifier is proposed to enhance impacts of positive labels.•Meta-paths are employed to capture implicit relations between labelers.

论文关键词：Homophily,Heterogeneous information network,Label correction,Crowdsourcing,Inference algorithm

论文评审过程：Received 11 May 2021, Revised 9 March 2022, Accepted 12 March 2022, Available online 26 March 2022, Version of Record 4 April 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116896