US 12,719,982 B2 · Forecasting call quality using machine learning techniques

Abstract

Methods, apparatus, and processor-readable storage media for forecasting call quality using machine learning techniques are provided herein. An example computer-implemented method includes obtaining call records for calls corresponding to a first time period, where each of the call records comprises a set of parameter values associated with the corresponding call. The method includes generating, using a first machine learning model, quality scores for the calls based on the corresponding call records, where the quality score for a given call is generated based on the set of parameter values in the corresponding call record. The method also includes generating, using a second machine learning model, a quality score forecast corresponding to a second time period based on the generated quality scores, identifying an issue affecting call quality during the second time period based on the quality score forecast, and initiating an automated action to at least partially remediate the identified issue.

Claims20

Dependent claims

A computer-implemented method comprising:

obtaining a plurality of call records for respective ones of a plurality of calls corresponding to a first time period, wherein the plurality of call records comprise respective sets of parameter values associated with the corresponding call, wherein the plurality of calls is implemented using at least one hardware-based communication device;
generating, using a first machine learning model, quality scores for the plurality of calls based at least in part on the plurality of corresponding call records, wherein the quality score for a given one of the plurality of calls is generated based on at least a portion of the set of parameter values in the corresponding call record;
generating, using a second machine learning model, a quality score forecast corresponding to a second time period based at least in part on a time series of the generated quality scores;
identifying at least one issue affecting a call quality during at least a portion of the second time period based at least in part on the quality score forecast;
and
initiating one or more automated actions to at least partially address the identified at least one issue;
wherein the method is performed by at least one processing device comprising a processor coupled to a memory.

4 more elements

A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device:

to obtain a plurality of call records for respective ones of a plurality of calls corresponding to a first time period, wherein the plurality of call records comprise respective sets of parameter values associated with the corresponding call, wherein the plurality of calls is implemented using at least one hardware-based communication device;
to generate, using a first machine learning model, quality scores for the plurality of calls based at least in part on the plurality of corresponding call records, wherein the quality score for a given one of the plurality of calls is generated based on at least a portion of the set of parameter values in the corresponding call record;
to generate, using a second machine learning model, a quality score forecast corresponding to a second time period based at least in part on a time series of the generated quality scores;
to identify at least one issue affecting a call quality during at least a portion of the second time period based at least in part on the quality score forecast;
and
to initiate one or more automated actions to at least partially address the identified at least one issue.

3 more elements

An apparatus comprising:

at least one processing device comprising a processor coupled to a memory;
the at least one processing device being configured:
to obtain a plurality of call records for respective ones of a plurality of calls corresponding to a first time period, wherein the plurality of call records comprise respective sets of parameter values associated with the corresponding call, wherein the plurality of calls is implemented using at least one hardware-based communication device;
to generate, using a first machine learning model, quality scores for the plurality of calls based at least in part on the plurality of corresponding call records, wherein the quality score for a given one of the plurality of calls is generated based on at least a portion of the set of parameter values in the corresponding call record;
to generate, using a second machine learning model, a quality score forecast corresponding to a second time period based at least in part on a time series of the generated quality scores;
to identify at least one issue affecting a call quality during at least a portion of the second time period based at least in part on the quality score forecast;
and
to initiate one or more automated actions to at least partially address the identified at least one issue.

5 more elements

Dependencies are inferred from the claim text.

Forecasting call quality using machine learning techniques

Life of this patent

Forecasting call quality using machine learning techniques

Life of this patent

Abstract

Claims20