Processing unstructured voice of customer feedback for improving content rankings in customer support systems

Patent metadata

Field	Value
Publication number	US-11734330-B2
Application number	US-202016800159-A
Country	US
Kind code	B2
Filing date	Feb 25, 2020
Priority date	Apr 8, 2016
Publication date	Aug 22, 2023
Grant date	Aug 22, 2023

Abstract

Techniques are disclosed for adjusting a ranking of information content presented to a user based on voice-of-customer feedback. In one embodiment, a user may provide feedback on information content presented to the user. Such feedback may be evaluated to identify at least one topic referenced in the received feedback. If an application determines that the at least one topic is related to topics of the information content, the application determines sentiment regarding the information content based on the feedback, and adjusts a ranking of the information content based on the determined sentiment.

First claim

What is claimed is:

1. A computer-implemented method for adjusting a ranking of information content of a software application having a content repository associated therewith and presented to a user based on voice-of-customer feedback, comprising:
receiving feedback regarding a content item selected from the content repository and presented to a user, the feedback including unstructured text content and structured feedback, the unstructured text content having at least one topic, each topic of the unstructured text content having a respective content type, and the structured feedback having a topic corresponding to a topic of the content item and a content type corresponding to a content type of the content item;
identifying, based on a probabilistic topic model generated from the content repository, at least one topic referenced in the unstructured text content;
determining, for each respective identified topic of the at least one identified topic, using metadata associated with the respective identified topic, a content type associated with the respective identified topic;
determining, for each respective identified topic of the at least one identified topic, whether the respective identified topic is related to a quality of the content item and whether the content type of the respective identified topic matches the content type of the content item;
for each identified topic of the at least one identified topic for which is it determined that the identified topic is related to the quality of the content item and that the content type of the identified topic matches the content type of the content item, applying a first weight to the structured feedback;
for each identified topic of the at least one identified topic for which it is determined that the identified topic is not related to the quality of the content item or that the content type of the identified topic does not match the content type of the content item, applying a weight of a one or more second weights having a value less than the first weight to the structured feedback;
ranking the content item based on the structured feedback, the first weight, and the one or more second weights; and
presenting on a user interface of another user a selection of search results that include the content item, wherein a position of the content item within the selection of search results is based on the ranking of the content item.

2. The computer-implemented method of claim 1 , wherein the structured feedback provides an indication of an experience of the user relative to the content item.

3. The computer-implemented method of claim 2 , wherein the structured feedback comprises either an up vote or a down vote or a star rating that indicates whether the user approves or disapproves of the content item.

4. The computer-implemented method of claim 1 , wherein the unstructured text content is evaluated using natural language processing.

5. The computer-implemented method of claim 1 , further comprising:
evaluating feedback from a plurality of other users regarding content items associated with a first topic of the at least one topic referenced in the unstructured text content;
identifying the first topic of the at least one topic as a trending topic based on the feedback from the plurality of other users;
selecting content items regarding the trending topic; and
presenting the selected content items regarding the trending topic to users.

6. The computer-implemented method of claim 1 , wherein the probabilistic topic model is a Latent Dirichlet Allocation (LDA) model or a correlated topics model (CTM).

7. The computer-implemented method of claim 1 , wherein the respective content type of each topic of the unstructured text content and the content type of the content item are selected from a group comprising: self-support content, assisted support content, product related content, application content, and service content.

8. A system, comprising:
one or more processors; and
a memory comprising instructions that, when executed on the one or more processors, cause the system to perform a method for adjusting a ranking of information content of a software application having a content repository associated therewith and presented to a user based on voice-of-customer feedback, the method comprising:
receiving feedback regarding a content item selected from the content repository and presented to a user, the feedback including unstructured text content and structured feedback, the unstructured text content having at least one topic, each topic of the unstructured text content having a respective content type, and the structured feedback having a topic corresponding to a topic of the content item and a content type corresponding to a content type of the content item;

identifying, based on a probabilistic topic model generated from the content repository, at least one topic referenced in the unstructured text content;
determining, for each respective identified topic of the at least one identified topic, using metadata associated with the respective identified topic, a content type associated with the respective identified topic;
determining, for each respective identified topic of the at least one identified topic, whether the respective identified topic is related to a quality of the content item and whether the content type of the respective identified topic matches the content type of the content item;
for each identified topic of the at least one identified topic for which is it determined that the identified topic is related to the quality of the content item and that the content type of the identified topic matches the content type of the content item, applying a first weight to the structured feedback;
for each identified topic of the at least one identified topic for which it is determined that the identified topic is not related to the quality of the content item or that the content type of the identified topic does not match the content type of the content item, applying a weight of a one or more second weights having a value less than the first weight to the structured feedback;
for each identified topic for which it is determined that the content type of the identified topic
ranking the content item based on the structured feedback, the first weight, and the one or more second weights; and
presenting on a user interface of another user a selection of search results that include the content item, wherein a position of the content item within the selection of search results is based on the ranking of the content item.

9. The system of claim 8 , wherein the structured feedback provides an indication of an experience of the user relative to the content item.

10. The system of claim 9 , wherein the structured feedback comprises either an up vote or a down vote or a star rating that indicates whether the user approves or disapproves of the content item.

11. The system of claim 8 , wherein the unstructured text content is evaluated using natural language processing.

12. The system of claim 8 , wherein the method further comprises:
evaluating feedback from a plurality of other users regarding content items associated with a first topic of the at least one topic referenced in the unstructured text content;
identifying the first topic of the at least one topic as a trending topic based on the feedback from the plurality of other users;
selecting content items regarding the trending topic; and
presenting the selected content items regarding the trending topic to users.

13. The system of claim 8 , wherein the probabilistic topic model is a Latent Dirichlet Allocation (LDA) model or a correlated

Assignees

• Intuit Inc

Inventors

• Podgorny Igor A
• Bartolome Warren
• Hung Kelvin
• Indyk Benjamin

Classifications

• G06F16/353 Primary

  into predefined classes · CPC title

• G06F16/337

  Profile generation, learning or modification · CPC title

• G06F16/3344

  using natural language analysis · CPC title

• G06F16/35 Primary

  Clustering; Classification · CPC title