Automatic labeling of river restoration project documents based on project objectives and restoration methods

作者：

Highlights：

• Machine-learning methods are developed to label river restoration project documents.

• A three-metric method is developed to determine majority labels in dependent MLC.

• An iterative data balancing method is developed to enhance learning-based MLC.

• Experiments from 1400 projects show satisfactory performance of the methods.

摘要

•Machine-learning methods are developed to label river restoration project documents.•A three-metric method is developed to determine majority labels in dependent MLC.•An iterative data balancing method is developed to enhance learning-based MLC.•Experiments from 1400 projects show satisfactory performance of the methods.

论文关键词：River restoration,Project document,Object labeling,Method labeling,Multilabel text classification

论文评审过程：Received 29 May 2021, Revised 7 September 2021, Accepted 24 February 2022, Available online 1 March 2022, Version of Record 3 March 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116754