SYSTEM AND METHOD FOR PERFORMING SPEECH ANALYTICS

• US 20120084081A1
• Filed: 09/30/2010
• Published: 04/05/2012
• Est. Priority Date: 09/30/2010
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Gilbert, Mazin, MELAMED, ILYA Dan
• Granted Patent: US 9,213,978 B2
• Time in Patent Office: Days
• US Class Current: 704/211
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 30/01 Customer relationship services

Abstract

Disclosed herein are systems, methods, and non-transitory computer-readable storage media for performing trend analysis of speech. A system practicing the method receives a speech trend analysis request having candidate feature constraints, an objective function with respect to a speech trend to be analyzed, and a set of speech record constraints. The system selects a subset of speech records from the group of speech records based on the set of speech record constraints to yield selected speech records, identifies features in the selected speech records based on the set of candidate feature constraints to yield identified features, and assigns a weight to each of the identified features based on the objective function. Then the system ranks the identified features by their respective weights to yield ranked identified features, and outputs at least one of the ranked identified features associated with a speech-based trend in response to the speech trend analysis request.

Claims

1. A method of performing trend analysis of speech, the method comprising:

receiving, as part of a speech trend analysis request, a set of candidate feature constraints, an objective function with respect to a speech trend to be analyzed, and a set of speech record constraints to be applied to a group of speech records;

selecting a subset of speech records from the group of speech records based on the set of speech record constraints to yield selected speech records;

identifying features in the selected speech records based on the set of candidate feature constraints to yield identified features;

assigning a weight to each of the identified features based on the objective function;

ranking the identified features by their respective weights to yield ranked identified features; and

outputting at least one of the ranked identified features associated with a speech-based trend in response to the speech trend analysis request.

2. The method of claim 1, wherein the group of speech records comprises at least one of recorded speech, transcriptions, and metadata.

3. The method of claim 1, wherein selecting the subset of speech records from the group of speech records further comprises:

generating a database query command corresponding to the speech record constraints;

executing the database query command.

4. The method of claim 3, wherein results of executing the database query command are passed to a feature filtering module which identifies the features in the selected speech records, and wherein the results of executing the database query are not displayed to a user.

5. The method of claim 1, wherein the set of candidate feature constraints constrains at least one of transcript features, metadata features, and higher order features inferred from the transcript features and metadata features.

6. The method of claim 1, wherein the set of candidate feature constraints is an empty set.

7. The method of claim 1, wherein the objective function is at least one of a mathematical function, a single feature, a combination of features, and time.

8. The method of claim 1, wherein the at least one of the ranked identified features is outputted via a user interface in which a user can focus on records with highly predictive features.

9. The method of claim 1, wherein, after outputting at least one of the ranked identified features, a user can revise the speech trend analysis request.

10. The method of claim 9, wherein revising the speech trend analysis request further comprises filtering out one or more feature constraint from the set of candidate feature constraints.

11. A system for generating an alert based on speech analytics data, the system comprising:

a processor;

a first module configured to control the processor to generate elements of a time series, wherein each element comprises speech records having timestamps within a same time interval;

a second module configured to control the processor to generate a numeric value for each element in the time series based on a weight for each speech record, wherein the weight is based on a set of candidate feature constraints, an objective function with respect to a trend to be analyzed, and a set of record constraints to be applied to a group of records; and

a third module configured to control the processor to generate an alarm when at least one respective numeric value for at least one element in the time series meets a threshold.

12. The system of claim 11, wherein a user specifies a size of the time interval.

13. The system of claim 11, wherein elements of the discrete time series are separated by an interval offset.

14. The system of claim 11, wherein the numeric value for each element in the time series is generated by summing respective scores for a plurality of features.

15. The system of claim 11, wherein the numeric value for each element in the time series is generated by averaging respective scores for a plurality of features.

16. The system of claim 11, wherein the third module is further configured to control the processor to notify a user when a specific combination of alarms is generated.

17. The system of claim 11, wherein the threshold is relative to other elements in the time series.

18. The system of claim 11, wherein the threshold is an absolute threshold.

19. A non-transitory computer-readable storage medium storing instructions which, when executed by a computing device, cause the computing device to perform trend analysis of speech, the instructions comprising:

receiving, as part of a speech trend analysis request, a set of candidate feature constraints, an objective function with respect to a speech trend to be analyzed, and a set of speech record constraints to be applied to a group of speech records;

selecting a subset of speech records from the group of speech records based on the set of speech record constraints to yield selected speech records;

identifying features in the selected speech records based on the set of candidate feature constraints to yield identified features;

assigning a weight to each of the identified features based on the objective function;

ranking the identified features by their respective weights to yield ranked identified features; and

outputting at least one of the ranked identified features associated with a speech-based trend in response to the speech trend analysis request.

20. The non-transitory computer-readable storage medium of claim 19, wherein, after outputting at least one of the ranked identified features, a user can revise the speech trend analysis request.