Interaction Design Group(7)

Following up with the speech technology development, spoken dialogue systems are getting commercialized. Many public transportation companies, call centres, tourist information centres, etc are using spoken dialogue systems instead of the touch-tone teleco

perform. But after a couple of times using the system, they became irritated and frustrated, either by the system’s unchanged boring voice or by the endless repeating of the long sentence all the time. The dialogue is not elegant and graceful at all.

The voice of SJ:s dialogue system lacks necessary intonation to certain sentences, which makes the user confused. For example, when the system does not recognize the subject’s sentence, the system says “what you sad was not understood”. This is said with a very formal voice, in a boring and monotone tone, without any personality or sympathy. In normal human-to human conversation, if one person does not understand the other person’s talk, this person would say: “excuse me, what did you say”, or “sorry, I did not hear”. Given the right prosody and intonation in this sentence it can provide a much more comforting message and give a very different feeling to the user. Subjects complained that the SJ:s dialogue system did not show any flattery and praise in the dialog. If the spoken language from the system contains more humanized attributes and personality, the subject would feel more natural in the conversation with the system. Not any interjection word was used in the dialogue, such as “oh” or “great” which made the dialogue feel stale and not alive.

Subjects did not get any personal contact or chemistry with the system. The spoken voice from the system is quite unpleasant and it affects the experience of interacting with it.

After finishing the experiment, two subjects said that they would not use this system again. The others gave positive comments about the system, but they also said that they still prefer to speak to real human operator even if they would have to wait in a queue.

5. Reversed Persona

All six subjects gave the same picture as what they gave before when regarding this company as a person. The only difference is that it became a man in the middle age as the voice of the system is male.

4. Discussion

We could not find any information about whether the company who designed this dialogue system, or SJ company itself have done any usability evaluation before. Besides, the understanding of the usability of a dialogue system can be very different. For many technical driving companies, the test of the functionality of the system is regarded as the usability tests [6]. If the system provides all the functions that are needed, then the system reaches its usability. The users’ experience of using the system, whether they like or dislike the system has not been considered for most of the spoken dialogue system designers.

In ISO 9241-11, the user’s satisfaction is important part of the usability. User’s satisfaction is a subjective issue, and it depends on the individual person, the preference of this person, the experience, the expectation and the context [9]. In this study, the three scenarios we designed are based on the most common tasks that users would like to perform with the system. In this experiment, all six subjects managed to go through all three scenarios and none of them gave way in the middle of the experiment. All of them learned how to adapt their speech behavior in order to make the system understand them. These results indicated that the tested dialogue system fulfilled the functional requirements. But the users’

搜索“diyifanwen.net”或“第一范文网”即可找到本站免费阅读全部范文。收藏本站方便下次阅读，第一范文网，提供最新人文社科Interaction Design Group(7)全文阅读和word下载服务。