基于过程识别的认知性VDT监控作业绩效形成机理  

A Mechanism Study of Cognitive VDT Monitoring Performance Based on Process Recognition

作　　者：黄琨 廖斌[2] HUANG Kun;LIAO Bin(Information Technology Department,Urban Vocational College of Sichuan,Chengdu 610101,China;School of Business,Sichuan Normal University,Chengdu 610101,China)

机构地区：[1]四川城市职业学院信息技术学院,四川成都610101 [2]四川师范大学商学院,四川成都610101

出　　处：《工业工程》2023年第4期27-34,43,共9页Industrial Engineering Journal

基　　金：教育部人文社会科学研究规划基金资助项目(19YJAZH051);四川省软科学研究资助项目(2020JDR0267)。

摘　　要：为提高认知性VDT监控作业绩效,运用隐马尔可夫模型解析作业过程并分析绩效形成机理.构建认知性VDT监控作业HMM概念模型,运用E-prime设计实验任务,运行ErgoLAB实验平台采集被试绩效及眼动数据,训练HMM参数并进行可信性验证,使用Viterbi算法将观察序列解码为认知动素链,分析认知动素链特征与作业绩效之间的关系,探究绩效形成机理.结果表明,认知性VDT监控作业过程可以用认知动素链表征,作业者和任务的不同会导致认知动素链的差异;结构相同情况下,认知动素链越长作业绩效越差;长度相同情况下,认知动素链结构和动素类型的差异会影响作业绩效.To improving the performance of cognitive visual display terminal(VDT)monitoring,the hidden Markov model(HMM)is used to analyze the operation process and the mechanism of performance formation.An HMM conceptual model of cognitive VDT monitoring task is established and the experiments are designed by E-Prime.Then,the performance and eye movement data are collected by ErgoLAB platform to train and test the HMM parameters.The observation sequences are decoded into cognitive therbligs chains by Viterbi algorithm.Finally,the relationship between the characteristics of cognitive therbligs chains and operation performance to explore the mechanism of performance formation.Results show that the process of cognitive VDT monitoring can be represented by cognitive therbligs chains,and the differences in operators and operations may lead to different chains;under the same structure,the operation performance decreases as the cognitive therbligs chain becomes longer;differences in structures and types of cognitive therbligs chains affect the operation performance when the lengths of chains are equal.

关 键 词：认知性VDT监控作业 认知动素链 隐马尔可夫模型 聚类分析 绩效机理 

分 类 号：X913.4[环境科学与工程—安全科学] TP399[自动化与计算机技术—计算机应用技术]