Method and system of optimal selection strategy for statistical classifications in dialog systems
First Claim
Patent Images
1. A processor-based dialog system comprising:
- a central processing unit;
a classifier circuit coupled to the central processing unit and configured to generate a plurality of predictions on a dialog act corresponding to an utterance received at an input; and
a decision circuit coupled to the central processing unit and configured to generate a plurality of probabilities each of which corresponds to a prediction, dynamically select a set of predictions from the plurality of predictions, and classify the utterance according to an accumulated probability of the set of predictions, wherein the accumulated probability is an accumulated probability of variable n-best outputs, wherein the value n is a non-fixed integer value that is determined dynamically based on a classifier distribution of the utterance.
1 Assignment
0 Petitions
Accused Products
Abstract
An optimal selection or decision strategy is described through an example that includes use in dialog systems. The selection strategy or method includes receiving multiple predictions and multiple probabilities. The received predictions predict the content of a received input and each of the probabilities corresponds to one of the predictions. In an example dialog system, the received input includes an utterance. The selection method includes dynamically selecting a set of predictions from the received predictions by generating ranked predictions. The ranked predictions are generated by ordering the plurality of predictions according to descending probability.
-
Citations
23 Claims
-
1. A processor-based dialog system comprising:
-
a central processing unit; a classifier circuit coupled to the central processing unit and configured to generate a plurality of predictions on a dialog act corresponding to an utterance received at an input; and a decision circuit coupled to the central processing unit and configured to generate a plurality of probabilities each of which corresponds to a prediction, dynamically select a set of predictions from the plurality of predictions, and classify the utterance according to an accumulated probability of the set of predictions, wherein the accumulated probability is an accumulated probability of variable n-best outputs, wherein the value n is a non-fixed integer value that is determined dynamically based on a classifier distribution of the utterance. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A computer-implemented method of classifying an utterance comprising:
-
receiving, in a classifier circuit of a processor-based dialog system, the utterance; generating, in a decision circuit of the processor-based dialog system, a plurality of predictions on a dialog act corresponding to the utterance; generating a plurality of probabilities each of which corresponds to a prediction; dynamically selecting a set of predictions from the plurality of predictions; and classifying, in a classifier circuit of the processor-based dialog system, the utterance according to an accumulated probability of the set of predictions, wherein the accumulated probability is an accumulated probability of variable n-best outputs, wherein the value n is a non-fixed integer value that is determined dynamically based on a classifier distribution of the utterance. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18, 19)
-
-
20. A physical, non-transitory computer readable media including executable instructions which, when executed in a processor-based dialog system, classify an utterance by:
-
receiving, in a classifier circuit of the processor-based dialog system, the utterance; generating, in a decision circuit of the processor-based dialog system, a plurality of predictions on a dialog act corresponding to the utterance; generating a plurality of probabilities each of which corresponds to a prediction; dynamically selecting a set of predictions from the plurality of predictions; and classifying, in a classifier circuit of the processor-based dialog system, the utterance according to an accumulated probability of the set of predictions, wherein the accumulated probability is an accumulated probability of variable n-best outputs, wherein the value n is a non-fixed integer value that is determined dynamically based on a classifier distribution of the utterance; and quantifying, in a performance measurement circuit coupled to the decision circuit, the performance of the decision circuit in classifying the utterance by deriving a Receiver Operating Curve (ROC) that graphs a false positive rate and a true positive rate, by varying the threshold, and calculating an area defined by the ROC. - View Dependent Claims (21, 22, 23)
-
Specification