摘要
患者投诉是改善医疗服务管理的珍贵资源。针对患者投诉分析需要大量人力、时间成本以及分类存在主观影响的问题,使用一种科学的分类标准对收集到的患者投诉进行人工标注,设计了一种基于长短期记忆模型的多个二元分类器结合的方法对患者投诉语料进行分类,探索了分类器对于单标签语料、多标签语料和仿真语料的分类预测的性能,为患者投诉分类提供了一种全新高效的方法,为更好地理解患者投诉打下坚实基础。
Patient complaints are a valuable resource for improving medical service management.A LSTM model-based automatic classification method of patient complaints with multiple binary classifiers was thus designed in accordance with the vast amount of manpower,time,costs and subjective influencing factors required for the analysis of patient complaints,and was used in classification of the words and phrases in patient complaints.The performance of classifiers was studied in classification of single-labeled words and phrases,multi-labeled words and phrases,and simulated words and phrases in order to provide a brand-new and effective method for the classification of patient complaints and to lay a solid foundation for the better understanding of patient complaints.
作者
姜垚松
马敬东
赵冬
罗玮
倪维斌
夏晨曦
JIANG Yao-song;MA Jing-dong;ZHAO Dong;LUO Wei;NI Wei-bin;XIA Chen-xi(Central China University of Science and Technology Tongji Medical College Medical and Health Management School,Wuhan 430030,Hubei Province,China)
出处
《中华医学图书情报杂志》
CAS
2018年第6期16-21,共6页
Chinese Journal of Medical Library and Information Science
基金
中央高校基本科研业务费资助华中科技大学自主创新基金项目"面向社交网络的情感分析与观点挖掘方法研究"(0118516036)
关键词
患者投诉
自动分类
长短期记忆模型
人工标注
标签
Patient complaints
Automatic classification
LSTM model
Artificial annotation
Tags
