AwareComm: Context Aware Communication for the Workplace

Project Team

Co- Directors: Melody Moore Jackson, PhD and Carrie Bruce, MA, CCC-SLP, ATP
Project Team: Jeremy Johnson, Georgia Tech
J. Higginbotham, PhD., University of Buffalo
Project Partner: DynaVox Technologies
Advisor: Kevin Caves, Duke University

Summary / Outcome Goals

Slow communication rate with existing AAC devices is one of the greatest challenges to their use. This problem is even more pronounced in situations, such as the workplace, where quick access to vocabulary is important. To address this problem, this project will partner with DynaVox Technologies to develop a context aware AAC device for office environments that will improve natural language processing. Additionally, we will develop context-specific office vocabulary for communication that can be used either in the new device or in other, existing AAC devices.

The specific aims of this project are to:

  1. characterize vocabulary relevant to verbal communication in office settings
  2. develop office-specific workplace vocabulary
  3. develop a context aware system that incorporates and utilizes information about time, location, and identity of a conversational partner to predict communication needs
  4. evaluate the new vocabulary and context aware technologies in DynaVox AAC devices
  5. transfer the context specific vocabulary and context aware technology to DynaVox for commercialization. Based on the results of this study, future projects will develop context-specific vocabulary for other work settings

This project is currently focused on two objectives: understanding workplace conversation and developing an utterance tagging application that will enable more personalized vocabulary storage and retrieval methods.

Workplace Conversation

A multi-phase formal study of workplace conversation across various communication channels including face-to-face, email, and instant messaging will be undertaken to investigate how AAC devices can be designed to better support workplace conversations. This study is currently investigating face-to-face task-oriented workplace discourse. We are working in coordination with faculty in the Department of Linguistics and ESL at Georgia State University to collect preliminary data. The activities of this project include:


We have developed the TagTalker application to provide an alternative method for storing and retrieving workplace relevant utterances. TagTalker uses the concept of tagging that has been made popular by sites such as Delicious and Flickr. These tags are non-hierarchical keywords or terms assigned to user-generated or imported utterances and can be added after a conversation has occurred or in anticipation of a future conversation.


This project has resulted in the following selected outputs:


Bruce, C. (2008). Critically Analyzing Workplace Discourse to Inform AAC Device Design. In proceedings of the 2nd Annual Clinical AAC Research Conference, Charlottesville, VA.


Bruce, C. (2008). Critically Analyzing Workplace Discourse to Inform AAC Device Design. Paper presented at the Doctoral Consortium at the 10th International ACM SIGAccess Conference on Computers and Accessibility, Halifax, NS, Canada.

Bruce, C., Moore-Jackson, M. and Johnson, J. (2008). AwareComm: Context aware communication for the workplace. Poster presented at the 13th Biennial Conference of the International Society for Augmentative and Alternative Communication, Montreal, QC, Canada

Ongoing Work

AwareComm is an ongoing project studying Alternative and Augmentative Communication (AAC) in a workplace environment. The driving goal of AwareComm is to narrow selection spaces for AAC users depending on context, to make conversational options more relevant to the current location, conversational partner, or situation. We have implemented "TagTalker", a communication system on the Dynavox platform, which allows words and phrases to be "tagged" with their context. This associative organization provides a much more efficient set of phrases to the AAC user. Our current work involves incorporating automated context sensing for location, to allow differentiation between conversation that might happen in a meeting room, an office, or a break room.