Conversational perspectives in a communication aid for non-vocal people

Area of Science:

• Augmentative and Alternative Communication (AAC)
• Human-Computer Interaction
• Speech Synthesis

Background:

• A computer-based text storage and retrieval system, TALK, was developed as a conversation aid for non-vocal individuals.
• TALK organizes potential utterances using three perspectives: person, time, and issue, to facilitate rapid retrieval and synthesized speech output.
• The uniform perspective organization may not be optimal for all conversational topics.

Purpose of the Study:

• To investigate whether topic-tailored perspective sets improve the speed and accuracy of storing and retrieving utterances compared to a uniform set.
• To evaluate the efficiency of different organizational structures for a conversational aid.

Main Methods:

• An experiment was conducted comparing a uniform set of perspectives with tailored sets for various conversational topics.
• Participants classified potential utterances for storage and retrieval under both organizational conditions.
• Performance was measured by the speed of classification during storage and accuracy during retrieval.

Main Results:

• The uniform set of perspectives facilitated faster classification during the storage stage.
• Tailored sets of perspectives resulted in more accurate retrieval decisions without increasing retrieval time.
• Accuracy in retrieval was higher with tailored perspectives, suggesting improved conversational flow.

Conclusions:

• While a uniform perspective set aids faster initial storage, topic-tailored perspectives enhance retrieval accuracy for conversational aids.
• Tailoring the organizational structure to specific topics may lead to more efficient and effective communication for non-vocal users.
• Further research into adaptive organizational strategies for AAC systems is warranted.