Configurable information collection system, method and computer program product utilizing speech recognition
View Patent ↗A system, method and computer program product are provided for collecting information utilizing speech recognition. After receiving a specification, a voice application is configured based on the specification. During use, the voice application is capable of collecting information utilizing speech recognition.
1. A method for collecting information utilizing speech recognition, comprising:
receiving a specification;
configuring a voice application based on the specification;
wherein the voice application is capable of collecting information utilizing speech recognition;
wherein the voice application is tested by collecting information utilizing the voice application, the information including answers to multiple choice questions, and confidence scores associated with answers to language queries.
2. A method for collecting information utilizing speech recognition, comprising:
receiving a specification;
configuring a voice application based on the specification;
wherein the voice application is capable of collecting information utilizing speech recognition;
wherein the voice application is tested by collecting information utilizing the voice application, the information including confidence scores associated with answers to language queries, the voice application being calibrated by manually setting at least one threshold for the confidence scores.
3. A method for collecting information utilizing speech recognition, comprising:
generating a mark-up language specification, the specification including a plurality of multiple choice questions, a plurality of language queries, and a plurality of open-ended questions;
configuring a voice application based on the specification;
testing the voice application by collecting information utilizing the voice application, the information including answers to the multiple choice questions, confidence scores associated with answers to the language queries, and recorded speech content received in response to the open-ended questions;
calibrating the voice application based on the testing, the calibration including manually setting at least one threshold for the confidence scores;
wherein the calibrated voice application is capable of collecting information utilizing speech recognition and providing a grade based on the at least one threshold, the confidence scores, and the answers to the multiple choice questions.