Granted Patent B2

US 12,505,289 · App. 18/429,781 · Granted Dec 23, 2025

Inventors: Matthew Dunn (Arlington, MA); Joe Bradley (Seattle, WI); Laura Onu (Redmond, WA)

Assignee: LIVEPERSON, INC.

G06F40/205G06F40/169G06N20/00

Abstract

Claims (64)

1 . A computer-implemented method comprising:

parsing a communication from an interaction involving a user device to identify one or more operative words from the communication;

determining an intent value selected from a list of intent categories associated with issues addressable by a communication system, wherein the intent value is associated with the one or more operative words in the communication, and wherein the intent value and the one or more operative words are associated with an action for the user device associated with the communication;

determining a confidence score for a strength of an association between the communication and the intent value;

determining that the confidence score is less than a threshold value;

determining a new intent value, wherein the new intent value is associated with the one or more operative words in the communication;

calculating a metric for the new intent value based on a quality of the association between the communication and the new intent value, wherein the metric is calculated using a correlation of one or more communications associated with the new intent value and one or more human agent profiles, wherein a human agent profile is associated with a human agent, and wherein the new intent value is associated with an additional action for the user device;

selecting a human agent profile based on the correlation; and

facilitating an additional interaction with the human agent associated with the human agent profile.

2 . The computer-implemented method of claim 1 , wherein determining the new intent value includes:

inputting the one or more operative words into a machine learning model trained with a training set of operative words and associated intent values; and

receiving, from the machine learning model, the new intent value.

3 . The computer-implemented method of claim 1 , wherein determining the new intent value includes:

using natural language data from previous communications.

4 . The computer-implemented method of claim 1 , wherein the list of intent categories is stored in a datastore, the method further comprising:

storing, in the datastore, the association of the new intent value with the one or more operative words.

5 . The computer-implemented method of claim 1 , further comprising matching the communication with a terminal device associated with the human agent profile, wherein matching is based on the metric.

6 . The computer-implemented method of claim 1 , wherein determining the new intent value includes:

presenting, in a user interface, the list of intent categories and the new intent value; and

receiving, through the user interface, a selection of the new intent value.

7 . The computer-implemented method of claim 6 , wherein presenting the new intent value includes presenting the new intent value as a subcategory of an intent category.

8 . A user device comprising:

a memory; and

one or more processors coupled to the memory, the one or more processors configured to perform operations comprising:

parsing a communication from an interaction involving a user device to identify one or more operative words from the communication;

determining an intent value selected from a list of intent categories associated with issues addressable by a communication system, wherein the intent value is associated with the one or more operative words in the communication, and wherein the intent value and the one or more operative words are associated with an action for the user device associated with the communication;

determining a confidence score for a strength of an association between the communication and the intent value;

determining that the confidence score is less than a threshold value;

determining a new intent value, wherein the new intent value is associated with the one or more operative words in the communication;

calculating a metric for the new intent value based on a quality of the association between the communication and the new intent value, wherein the metric is calculated using a correlation of one or more communications associated with the new intent value and one or more human agent profiles, wherein a human agent profile is associated with a human agent, and wherein the new intent value is associated with an additional action for the user device;

selecting a human agent profile based on the correlation; and

facilitating an additional interaction with the human agent associated with the human agent profile.

9 . The user device of claim 8 , wherein determining the new intent value includes:

inputting the one or more operative words into a machine learning model trained with a training set of operative words and associated intent values; and

receiving, from the machine learning model, the new intent value.

10 . The user device of claim 8 , wherein determining the new intent value includes:

using natural language data from previous communications.

11 . The user device of claim 8 , wherein the list of intent categories is stored in a datastore, the operations further comprising:

storing, in the datastore, the association of the new intent value with the one or more operative words.

12 . The user device of claim 8 , the operations further comprising: matching the communication with a terminal device associated with the human agent profile, wherein matching is based on the metric.

13 . The user device of claim 8 , wherein determining the new intent value includes:

presenting, in a user interface, the list of intent categories and the new intent value; and

receiving, through the user interface, a selection of the new intent value.

14 . The user device of claim 13 , wherein presenting the new intent value includes presenting the new intent value as a subcategory of an intent category.

15 . A non-transitory computer readable medium comprising instructions that, when executed by one or more processors of a user device, cause the user device to perform operations comprising:

parsing a communication from an interaction involving a user device to identify one or more operative words from the communication;

determining an intent value selected from a list of intent categories associated with issues addressable by a communication system, wherein the intent value is associated with the one or more operative words in the communication, and wherein the intent value and the one or more operative words are associated with an action for the user device associated with the communication;

determining a confidence score for a strength of an association between the communication and the intent value;

determining that the confidence score is less than a threshold value;

determining a new intent value, wherein the new intent value is associated with the one or more operative words in the communication;

calculating a metric for the new intent value based on a quality of the association between the communication and the new intent value, wherein the metric is calculated using a correlation of one or more communications associated with the new intent value and one or more human agent profiles, wherein a human agent profile is associated with a human agent, and wherein the new intent value is associated with an additional action for the user device;

selecting a human agent profile based on the correlation; and

facilitating an additional interaction with the human agent associated with the human agent profile.

16 . The non-transitory computer readable medium of claim 15 , wherein determining the new intent value includes:

inputting the one or more operative words into a machine learning model trained with a training set of operative words and associated intent values; and

receiving, from the machine learning model, the new intent value.

17 . The non-transitory computer readable medium of claim 15 , wherein determining the new intent value includes:

using natural language data from previous communications.

18 . The non-transitory computer readable medium of claim 15 , wherein determining the new intent value includes:

presenting, in a user interface, the list of intent categories and the new intent value; and

receiving, through the user interface, a selection of the new intent value.

19 . The non-transitory computer readable medium of claim 18 , wherein presenting the new intent value includes presenting the new intent value as a subcategory of an intent category.

20 . The non-transitory computer readable medium of claim 19 , wherein the list of intent categories is stored in a datastore, and wherein when executed by the one or more processors, the instructions cause the one or more processors to perform operations comprising:

storing, in the datastore, the association of the new intent value with the one or more operative words.

Assignments (5)

Continuity (5)

References Cited (12)