多通道界面软件结构模型及整合算法  

SOFTWARE MODEL AND INTEGRATION ALGORITHM OF MULTIMODAL INTERFACE

作　　者：李茂贞[1] 戴国忠[1] 董士海[2] 

机构地区：[1]中国科学院软件研究所,北京100080 [2]北京大学计算机科学与技术系,北京100871

出　　处：《计算机学报》1998年第2期111-118,共8页Chinese Journal of Computers

基　　金：国家自然科学基金!69433020

摘　　要：多通道界面是利用多个输入通道进行人机交互的新一代人机界面，旨在提高人机交互的自然性和高效性．本文首先阐述了多通道界面的概念和研究问题，然后提出评价多通道界面人机交互效率的一种机制．结合面向任务的观点提出一种多通道界面的软件结构模型，在此基础上实现了一个具有多通道界面特点的原型系统（MMI－PRO），并给出了相应的通道整合算法．最后利用提出的评价机制对MMI＿PRO交互效率进行了定量评价．Multimodal interface uses more than one input modalities to interactwith users to promote the nature and efficiency of human-computer interaction.Firstly, this paper describes the concepts and research issues of multimodalinterface. Secondly, It gives a mechanism of assessing the interaction efficiency ofmultimodal interface. Thirdly, a software structure model of multimodal interface,which is task-oriented, is presented. on the basis of the model,a prototype system(MMI-PRO), which is characteristic of multimodal interface, is implemented,andan integrated algorithm of multimodalities is also put forward. Finally, theinteraction efficiency of the MMI-PRO is evaluated in quantum using the assessingmechanism provided in this paper.

关 键 词：人机交互 多通道界面 多通道整合 软件结构模型 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]