I began developing this method of linguistic technology (technological applications of linguistic analysis) in the early to mid 1980's. For a description of my research, please see sections on methodology and publications. The early applications of linguistic technology were designed to be to be applied to computer-assisted clinical diagnostic and decision-making systems. The goal was to uncover hidden and obscure medical history data by identifying specific "sequence packages": lexical units and corresponding prosody as they are packaged within sequences of utterances. |
||||
By the process of identifying sequence packages located in a patient's dialogue, one is then able to uncover historical health data that may otherwise be unobtainable by virtue of their masked presentation. There are numerous causes for the obscurity of historical health data: Patients are frightened and apprehensive; physicians in their haste to make critical decisions can narrow the discourse opportunities for patients to fully present accurate historical information; and lastly, patients describe their symptoms anecdotally whereas physicians process symptom descriptions as conforming to classical textbook definitions. The Virtual Physician is of great benefit here. The sequence package approach to understanding medical history dialogue can be applied to a wide spectrum of applications. Just as a patient encounters difficulty in clearly articulating symptom descriptions, a consumer may also display ambiguity and circumlocution in articulating product descriptions. The frustrated consumer, who has a rough idea of a desired product but who cannot effectively communicate the product features to the customer service representative may likewise benefit from a sequence package analysis of his/her utterances. In this instance, the Automated Operator and Trouble-Shooter is quite useful. Lastly, the sequence package approach can be applied to both dictation programs, which convert continuous speech into written text, and letter writing wizards used to assist in textual construction by uncovering the interactional work that the user is attempting to accomplish. For example, if the user is trying to make a request in a deferential manner due to his subordinate position vis a vis the recipient of his request, but is wholly unclear and roundabout due to his preservation of a deferential tone, a sequence package approach can spot this request attempt made by the user. Alternate phraseology can then be supplied so as to preserve the deferential tone while making the message much clearer. The Automated Secretary assists with this. As a future project, LTS intends to develop a truth
verification device which can utilize the same sequence
package methodology by locating patterns within utterance
sequences indicative of lies and deceit, hence, the
Virtual Lie Detector. |
Selected Publications
|
|