Manage Your Agents: An Automatic Tool for Classification of Voice Intelligent Agents

Abstract

Voice intelligent agents absorb and deliver information among users, thus bringing information risks to users. This article designed an automatic tool called VIARS for the classification work of voice intelligent agents and compared the different management style of experts and general users for VIARS. Results found that experts relied less on VIARS, made more changes for VIARS recommendations, and achieved higher consistency in final classification results. In contrast, general users relied highly on VIARS recommendations and showed defensive tendency for classification recommendations. In addition, VIARS designed for expert are more useful in evaluating rule-based items while VIARS designed for general users are more useful in evaluating experience-based items.