A Study of User's Performance and Satisfaction on the Web Based Photo Annotation with Speech Interaction

TL;DR

This study empirically evaluates user performance and satisfaction with a web-based speech photo annotation system, showing that the proposed speech interaction significantly reduces annotation time and improves user satisfaction compared to Microsoft Vista speech style.

Contribution

It introduces a novel speech interaction style for web photo annotation and provides empirical evidence of its effectiveness over existing Microsoft Vista speech interaction.

Findings

Proposed speech interaction reduces annotation time by 64.72%.

Participants reported higher satisfaction with the proposed speech style.

Statistical analysis confirms significant improvements in performance and satisfaction.

Abstract

This paper reports on empirical evaluation study of users' performance and satisfaction with prototype of Web Based speech photo annotation with speech interaction. Participants involved consist of Johor Bahru citizens from various background. They have completed two parts of annotation task; part A involving PhotoASys; photo annotation system with proposed speech interaction and part B involving Microsoft Microsoft Vista Speech Interaction style. They have completed eight tasks for each part including system login and selection of album and photos. Users' performance was recorded using computer screen recording software. Data were captured on the task completion time and subjective satisfaction. Participants need to complete a questionnaire on the subjective satisfaction when the task was completed. The performance data show the comparison between proposed speech interaction and Microsoft Vista Speech interaction applied in photo annotation system, PhotoASys. On average, the reduction in annotation performance time due to using proposed speech interaction style was 64.72% rather than using speech interaction Microsoft Vista style. Data analysis were showed in different statistical significant in annotation performance and subjective satisfaction for both styles of interaction. These results could be used for the next design in related software which involves personal belonging management.