A Model of Persuasion for Speaking Rate Adaptation

Abstract

Proper speaking rate is a key attribute of effective communication. Emerging persuasive technologies use computers as a tool to induce human behavioural and attitude changes. This thesis established a computational framework which can persuade people to slow down their speech and communicate more effectively. We defined a conceptual model and implemented a computer software system, both serving as the cornerstones of our persuasion framework. The computer system is designed to persuade people to be aware of their speaking rate and to slow down their speech. The combination of computer technology and persuasive technologies and theories are embedded in the system. In order to conduct effective persuasion, a number of computer-based survey questions were asked and a short tailored letter was generated for each participant. A virtual coach system monitored and reminded the participant to slow down. A few adaptive cues were used to enhance the effects of the persuasion. We evaluated the feasibility and effectiveness of the overall system. At the same time, we evaluated the feasibility of individual elements. A total of 22 participants was selected to make up the sample. The experiments were conducted under controlled conditions. The results indicated that our system is effective in persuading people to speak more slowly. The feedback from users indicated that our system raised their awareness about speaking rate.