CONVERSATIONAL PERSPECTIVES IN A COMMUNICATION AID FOR NONVOCAL PEOPLE

A computer-based text storage and retrieval system (TALK) has recently been developed as a conversation aid for non-vocal people. To facilit ate the rapid retrieval (and synthesised speech output) of stored cont ent, potential utterances on each conversational topic are organised a ccording to three intersecting perspectives. The perspectives are 'per son' (me, you), 'time' (past, present, future), and 'issue' (what, whe n, where, who, how, why). Although the system has been shown capable o f supporting free-flowing conversation with relatively short pauses pr eceding the user's turns at speech, the perspective organisation is pr obably less than ideal for some topics. An experiment was conducted to see whether, over a number of conversational topics, storage and retr ieval of potential utterances would be faster with the existing, unifo rm, set of perspectives or with different sets of perspectives tailore d to each topic. At the storage stage, the 'uniform' set of perspectiv es facilitated more rapid classification of potential utterances than the 'tailored' sets. At the retrieval stage, however, repetition of th e classification decisions was more accurate (without being any slower ) for the tailored sets of perspectives, suggesting that retrieval of stored utterances during conversation may be faster when the perspecti ve organisation is tailored to topics.