Automated call classification

Patent metadata

Field	Value
Publication number	US-9961202-B2
Application number	US-201514985533-A
Country	US
Kind code	B2
Filing date	Dec 31, 2015
Priority date	Dec 31, 2015
Publication date	May 1, 2018
Grant date	May 1, 2018

Abstract

A computing system is configured to categorize digital data records each relating to an interaction between a call center agent and a customer. The system includes a rules module to determine a first probability that a data record belongs to a category based one or more rules each rule defining a relationship between one or more features of the data record and the category; a probabilistic module to determine a second probability that the data record belongs to the category based on one or more probabilistic models of relationships between one or more of the features and the category; an arbitration module to determine whether or not the data record belongs to the category based on the first and second probabilities; and an action module to receive from the arbitration processor an indication that the interaction belongs to the category and to initiate an action appropriate to the category.

First claim

What is claimed is:

1. A computing system configured to categorize digital data records each data record relating to an interaction between a call center agent and a customer, and each interaction being an exchange between parties including one or more of: an exchange via telephone, an exchange with an interactive voice response “IVR” system, a web chat, an exchange via instant messaging, an exchange via short message service “SMS” messages, and an exchange via email and each data record comprising at least one of: text, automatically generated codes and audio information, each data record having one or more features indicative of a category of the data record, the system comprising:
a memory; and
a processor comprising one or more processing modules including:
a rules module configured to receive as input a data record and to determine a first probability that the data record belongs to a category based one or more rules, each rule defining a relationship between at least one of said features and the category;
a probabilistic module configured to receive as input said data record and to determine a second probability that the data record belongs to the category based on one or more probabilistic models of relationships between one or more of said features and the category;
an arbitration module configured to determine whether or not the data record belongs to the category based on the first and second probabilities by converting the first probability into a first vote by comparing the first probability to a threshold and converting the second probability into a second vote by comparing the second probability to the threshold and if the first vote and second vote agree, determining whether the data record belongs to the category based on the votes, and if the votes disagree determining whether the data record belongs to the category based on the highest of the first probability and second probability weighted by historical performance; and
an action module configured to receive from the arbitration module an indication that the interaction belongs to the category and to initiate an action appropriate to the category, the action selected from the group consisting of: follow-up interaction between a call center agent and the customer; initiation of an interaction with an party other than the customer; setting of a flag to ensure that the call reason has been resolved; scheduling of a follow up interaction; triggering of a preemptive email concerning issues known to be associated with the category; and
automatic creation of customer history.

2. A system according to claim 1 in which the processing modules are configured to perform the determination whether or not a data record belongs to a category for each of a plurality of categories.

3. A system according to claim 1 wherein the processing modules further comprise a rule builder configured to use:
rules defining binary relationships between said at least one of said features and said category, and
a set of correctly classified data records,
to determine the first probability.

4. A system according to claim 3 in which the first probability is derived from the binary rules and the correctly classified data records using Bayes theory.

5. A system according to claim 3 in which the first probability is the sum of a calculated probability derived using the correctly classified data records and a biasing factor towards the upper confidence bound of the probability value.

6. A system according to claim 1 in which the probabilistic models are based on any of linear regression, logistic regression, decision tree, random forest, neural network, support vector machine and Bayesian network.

7. A system according to claim 1 in which the processing modules further comprise a probabilistic model builder configured to use:
features from the data records, and
a set of correctly classified records,
to derive at least one probabilistic model estimating the probability of a data record belonging to a category according to which of the extracted features are present in the data record;
and wherein the probabilistic models module is configured to use the derived probabilistic model(s) to determine the second probability.

8. A system according to claim 1 further comprising an arbitration strategy builder configured to determine a metric of the accuracy of the first and second probabilities respectively in determining whether a data record belongs to a category.

9. A system according to claim 8 wherein the arbitration strategy builder is further configured to perform the determination using a set of correctly classified data records and to repeat the determination using new correctly classified data records to update the metrics.

10. A system according to claim 1 wherein the first and second probabilities are weighted according to their distance from the predetermined threshold.

11. A method of categorizing digital data records, each data record relating to an interaction between a call center agent and a customer, each interaction being an exchange between parties including one or more of: an exchange via telephone, an exchange with an interactive voice response “IVR” system, a web chat, an exchange via instant messaging, an exchange via short message service “SMS” messages, and an exchange via email, each data record having one or more features and the method being carried out in a computer, the method comprising:
using a set of soft rules relating at least one of said features to a category to determine a first probability that a data record belongs to the category,
using a set of trained probabilistic models relating at least one of said features to said category to determine a second probability that a data record belongs to the category, converting the first probability into a first vote by comparing the first probability to a threshold and converting the second probability into a second vote by comparing the second probability to the threshold; and
determine on the basis of the first and second probabilities whether the data record belongs to a category by:
if the first vote and second vote agree, determining whether the data record belongs to the category based on the votes, and if the votes disagree determining that the data record belongs to the category based on the highest of the first probability and second probability weighted by historical performance.

12. A method according to claim 11 wherein at least one of the probabilistic models relates at least two different features to a category and is configured to take account of the relative importance of said different features.

13. A method according to claim 11 wherein the soft rules, the probabilistic models and the arbitration strategy are devised using a set of correctly classified data records, and where in the method further comprises updating one or more of the soft rules, the probabilistic models and the arbitration strategy using a new set of correctly classified data records.

14. A system for determining reasons for interactions between call center agents and customers, each interaction being represented as a data record having one or more features indicative of the reason for the call, and each interaction being an exchange between parties including one or more of: an exchange via telephone, an exchange with an interactive voice response “IVR” system, a web chat, an exchange via instant messaging, an exchange via short message service “SMS” messages, and an exchange via email, the system comprising:
a memory; and
a processor comprising one or more processing modules including:
a rules module configured to receive as

Assignees

• Nice Ltd

Inventors

• Bellosi Facundo
• Newnham Leonard Michael
• Ram Tomer

Classifications

• G06F17/30598

  Physics · mapped topic

• G06F17/30525

  Physics · mapped topic

• H04M3/5175 Primary

  Call or contact centers supervision arrangements · CPC title

• H04M2203/559

  Sorting systems · CPC title

• H04M2203/355

  Interactive dialogue design tools, features or methods · CPC title