Automatic classification of citizen requests for transportation using deep learning: Case study from Boston city

作者：

Highlights：

• Increasing volume of civil complaints necessitated automatic classification.

• Literature understudied a classification framework of unstructured real data.

• Five stepwise models were applied for the classification accuracy.

• SMOTE-ENN, balancing categories, significantly improved the accuracy.

摘要

•Increasing volume of civil complaints necessitated automatic classification.•Literature understudied a classification framework of unstructured real data.•Five stepwise models were applied for the classification accuracy.•SMOTE-ENN, balancing categories, significantly improved the accuracy.

论文关键词：Citizen requests,Unstructured data,Convolutional neural network,Imbalanced data

论文评审过程：Received 26 February 2020, Revised 30 August 2020, Accepted 9 October 2020, Available online 9 November 2020, Version of Record 9 November 2020.

论文官网地址：https://doi.org/10.1016/j.ipm.2020.102410