Design and Implementation of a Voice Interactive Tool to Facilitate Web Collaboration

This paper proposes a voice interactive tool that is based on a dual voice service model for Web-based collaboration. Instead of restricting team members to rely on the traditional interaction devices, such as keyboards, mice or display screens, this tool aims to enable each of them to "speak to the collaborative tool or the browser" and to obtain multisensory feedback, e.g., audiovisual or voice information. In that way, some specific collaborative tasks could be completed through natural language dialogue. The paper describes and highlights the significance and necessity of multimodal interaction for Web collaboration activities involving more and more smart devices, especially, how integrated voice interaction can provide a friendlier experience for teamwork. For some data-sensitive tasks, the local voice services are very beneficial and flexible because there is no need to upload the given set of voice data to a third party, and the related end-to-end speech models can be adjusted and trained in accordance with certain collaboration requirements. The voice supported collaborative editing and voice supported browsing have been tested as practical examples in a collaborative working environment MCWE2.0. The results show that the tool takes the advantages of voice interaction to expand the potential possible usage of multiple smart devices to a certain extent, thereby facilitating multimodal Web-based collaboration in a dual way.