Systems and methods for multi-channel customer communications content recommender

Patent metadata

Field	Value
Publication number	US-12493845-B2
Application number	US-202217959527-A
Country	US
Kind code	B2
Filing date	Oct 4, 2022
Priority date	Mar 3, 2020
Publication date	Dec 9, 2025
Grant date	Dec 9, 2025

Abstract

Interaction events collected across disparate customer communication channels of an enterprise are processed to generate an encoded unique content item identifier for each content item referenced in an interaction event such that the content item is resolvable to a location in a content repository. A training data set is built using the interaction events thus processed and a multi-channel content recommendation model is trained using the training data set. The multi-channel content recommendation model thus trained stores data points representing intersections of customers and content items that the enterprise has been tracking, with each data point having an effectiveness score for an associated content item. The multi-channel content recommendation model thus trained can be queried by content designers of the disparate customer communication channels through a recommender application for content recommendations based on the effectiveness of the content, agnostic to the disparate customer communication channels.

First claim

What is claimed is:

1 . A method, comprising:
feeding customer communications events from across a plurality of communication channels into an event pipeline, the customer communications events associated with a plurality of full communications across the plurality of communication channels, the plurality of full communications comprising communication assets, the event pipeline configured for processing the customer communications events into training data for training a recommendation engine;
generating a sparse matrix of effectiveness scores based on available interaction data, wherein the interaction data represents user interactions with content items;
transforming the sparse matrix into a populated matrix by executing a process on a processor, the process comprising:
determining an event-based data point within the sparse matrix, wherein the event-based data point is a sparse intersection of a customer communication event and the communication asset;
identifying an additional data point for the sparse matrix for which interaction data is not available;

computing an additional effectiveness score for the additional data point of the sparse matrix, wherein the additional effectiveness score is computed using at least one of a content-based filter or a collaborative filter;
filling the additional data point of the sparse matrix with the computed effectiveness score to generate a populated matrix;
generating a limited communication for a limited communication channel, the limited communication comprising non-interactive content;
selecting, based on the effectiveness scores of the populated matrix, at least one content item from the communication assets of the full communication; and
inserting the at least one content item into the limited communication, thereby transforming the limited communication into an additional full communication.

2 . The method according to claim 1 , wherein the plurality of communication channels comprises at least one of: a web channel, an email channel, a social media channel, or a messaging channel, and wherein the limited communication channel comprises a print channel.

3 . The method according to claim 1 , wherein the communication assets comprise an image and a link, wherein inserting at least one of the communication assets comprises inserting at least one of: the image or the link.

4 . The method according to claim 1 , wherein the recommendation engine generates an event model associated with a full communication, the event model comprising a hierarchical data structure of interaction information recorded through the event pipeline.

5 . The method according to claim 4 , wherein the recommendation engine modifies an instance of the event model based on a newly generated effectiveness score.

6 . The method according to claim 1 , further comprising:
collecting or generating the customer communications events across the plurality of communication channels.

7 . The method according to claim 1 , further comprising:
generating the plurality of full communications via input received from communication channel modelers.

8 . A system, comprising:
a processor; a non-transitory computer-readable medium; and instructions stored on the non-transitory computer-readable medium and translatable by the processor for: feeding customer communications events from across a plurality of communication channels into an event pipeline, the customer communications events associated with a plurality of full communications across the plurality of communication channels, the plurality of full communications comprising communication assets, the event pipeline configured for processing the customer communications events into training data for training a recommendation engine;
generating a sparse matrix of effectiveness scores based on available interaction data, wherein the interaction data represents user interactions with content items;
transforming the sparse matrix into a populated matrix by executing a process on a processor, the process comprising:
determining an event-based data point within the sparse matrix, wherein the event-based data point is a sparse intersection of a customer communication event and the communication asset;
identifying an additional data point for the sparse matrix for which interaction data is not available;

computing an additional effectiveness score for the additional data point of the sparse matrix, wherein the additional effectiveness score is computed using at least one of a content-based filter or a collaborative filter;
filling the additional data point of the sparse matrix with the computed effectiveness score to generate a populated matrix;
generating a limited communication for a limited communication channel, the limited communication comprising non-interactive content;
selecting, based on the effectiveness scores of the populated matrix, at least one content item from the communication assets of the full communication; and
inserting the at least one content item into the limited communication, thereby transforming the limited communication into an additional full communication.

9 . The system of claim 8 , wherein the plurality of communication channels comprises at least one of: a web channel, an email channel, a social media channel, or a messaging channel, and wherein the limited communication channel comprises a print channel.

10 . The system of claim 8 , wherein the communication assets comprise an image and a link, wherein inserting at least one of the communication assets comprises inserting at least one of: the image or the link.

11 . The system of claim 8 , wherein the recommendation engine generates an event model associated with a full communication, the event model comprising a hierarchical data structure of interaction information recorded through the event pipeline.

12 . The system of claim 11 , wherein the recommendation engine modifies an instance of the event model based on a newly generated effectiveness score.

13 . The system of claim 8 , wherein the instructions are further translatable by the processor for:
collecting or generating the customer communications events across the plurality of communication channels.

14 . The system of claim 8 , wherein the instructions are further translatable by the processor for:
generating the plurality of full communications via input received from communication channel modelers.

15 . A computer program product comprising a non-transitory computer-readable medium storing instructions translatable by a processor for:
feeding customer communications events from across a plurality of communication channels into an event pipeline, the customer communications events associated with a plurality of full communications across the plurality of communication channels, the plurality of full communications comprising communication assets, the event pipeline configured for processing the customer communications events into training data for training a recommendation engine;
generating a sparse matrix of effectiveness scores based on available interaction data, wherein the interaction data represents user interactions with content items;
transforming the sparse matrix into a populated matrix by executing a process on a processor, the process comprising:
determining an event-based data point within the sparse matrix, wherein the event-based data point is a sparse intersection of a customer communication event and the communication asset;
identifying an additional data point for the sparse matrix for which interaction data is not available;

computing an additi

Assignees

• Open Text Sa Ulc

Inventors

• Pruitt Byron Steven

Classifications

• G06Q30/0204

  Market segmentation · CPC title

• G06Q30/0201

  Market modelling; Market analysis; Collecting market data · CPC title

• G06Q30/0281 Primary

  Customer communication at a business location, e.g. providing product or service information, consulting · CPC title

• G06Q10/0637 Primary

  Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals · CPC title