MODELING HUMAN TRANSCRIPTION TYPING WITH QUEUING NETWORK-MODEL HUMAN PROCESSOR (QN-MHP)

摘要

Typing is one of the basic and prevalent activities in human machine interaction. John (1988, 1996) proposed a PERT (Project-Evaluation-Research-Technique)-based model called TYPIST, which modeled 21 of the 31 behavioral phenomena in transcription typing (Salthouse, 1986, 1987; Gentner, 1983). However, TYPIST can only analyze the typing phenomena along the time dimension; it can not model error and eye movement of typing. Based on the queuing network theory of human performance (Liu, 1996, 1997) and current discoveries in brain and cognitive sciences, this paper proposes a queuing network model of typing which successfully modeled not only all the 21 phenomena modeled by TYPIST, but also 13 additional phenomena in transcription typing including 5 typing error phenomena, 3 eye movement phenomena and 2 brain imaging phenomena. Further developments of the queuing network model in modeling typing and other tasks, and its value in proactive ergonomic design of typing interfaces are discussed.