An exploratory study of the impact of training times on user acceptance of speech recognition systems and its managerial implications

摘要

Speech Recognition Systems (SRS) offer many possibilities with their ability to understand voice input and convert that input into text that the computer can understand and process. For example, SR systems may provide a more natural interface with current Group Decision Support Systems to enhance user satisfaction and text generation. However, one of the limitations of current SR systems is a training requirement. Most SR systems are 'speaker-dependent', which requires users to spend time training the computer to understand their voice. This paper reports the results of an exploratory study consisting of two experiments to determine the type of relationship that exists between training time, accuracy and end-user acceptance of SR technology. Subjects that gained more experience with the software through increased training time reported higher accuracy rates and generated more text in the same amount of time. In addition, they expressed a stronger preference for adoption of the SR technology. These results provide some insight towards the minimum training time required for successful adoption and user satisfaction of SR technology.