Granted Patent B2

US 12,314,542 · App. 18/057,932 · Granted May 27, 2025

Inventors: Phu Pham (Lawrenceville, GA); Merle Hidinger (Midlothian, VA); Jun Ji (Glen Allen, VA)

Assignee: TRUIST BANK

G06F3/0484G06F40/30G06N3/0442G06N3/045

Abstract

Claims (41)

1. A system for processing interactive content comprising a network computing device, wherein the network computing device comprises one or more integrated software applications that perform the operations comprising:

(a) receiving by the network computing device, content data files that each comprise (i) a plurality of communication elements, and (ii) sequencing data;

(b) executing a concentration analysis using the content data files, wherein the concentration analysis generates concentrated content data for each content data file by performing the operations comprising

(i) determining a weight quantifier for the communication elements, and

(ii) removing from the content data file, communication elements having a weight quantifier below a weight threshold;

(c) executing a subject classification analysis using the concentrated content data, wherein the subject classification analysis processes the concentrated content data for each of the content data files to generate

(i) one or more subject identifiers, and

(ii) subject weighting data for each subject identifier;

(d) executing a sentiment analysis using the concentrated content data, wherein the sentiment analysis generates a sentiment identifier for each of the content data files;

(e) aggregating, across all of the content data files processed by the subject classification analysis, the one or more subject identifiers, and generating subject proportion data for each distinct subject identifier;

(f) aggregating, across all of the content data files processed by the subject classification analysis, the one or more sentiment identifiers, and generating sentiment proportion data for each distinct sentiment identifier; and

(g) receiving a layout display command comprising layout display selection data transmitted from an agent computing device; and

(h) processing the layout display command by performing the operations comprising

(i) when the layout display selection data comprises a subject layout selection, transmitting the subject identifiers, the subject weighting data, subject proportion data for a subject threshold number of distinct subject identifiers, and the sequencing data to the agent user computing device for display, and

(ii) when the layout display selection data comprises a sentiment layout selection, transmitting sentiment identifiers, sentiment proportion data for a sentiment threshold number of distinct subject identifiers, and the sequencing data to the agent user computing device for display.

2. The system for processing interactive content of claim 1 , wherein the network computing device performs the further operations comprising:

(a) receiving content parameter data comprising one or more sequencing identifiers that each represent a sequence range;

(b) determining whether each of the content data files falls within a sequencing range by processing the content data files and the sequencing data for each of the content data files;

(c) labeling the content data files with at least one of the sequencing identifiers when the sequencing data falls within at least one of the sequencing ranges;

(d) aggregating, across all of the labeled content data files, the one or more subject identifiers, and generating subject proportion data for each distinct subject identifier;

(e) aggregating, across all of the labeled content data files, the one or more sentiment identifiers, and generating sentiment proportion data for each distinct sentiment identifier; and

(f) processing the layout display command by performing the further operations comprising

(i) when the layout display selection data comprises the subject layout selection, transmitting the subject proportion data for a subject threshold number of distinct subject identifiers to the agent user computing device for display, and

(ii) when the layout display selection data comprises the sentiment layout selection, transmitting sentiment proportion data for a sentiment threshold number of distinct subject identifiers to the agent user computing device for display.

3. The system for processing interactive content of claim 1 , wherein:

(a) the network computing device comprises a neural network; and

(b) the neural network is used to execute the subject classification analysis.

4. The system for processing interactive content of claim 3 , wherein the neural network performs operations that implement a Kmeans clustering analysis to execute the subject classification analysis.

5. The system for processing interactive content of claim 4 , wherein the neural network comprises a neural network architecture selected from one of a Hopefield network, a Boltzmann Machine, Sigmoid Belief Net, a Deep Belief Network, a Helmholtz Machine, a Kohonen Network, a Self-Organizing Map, or a Centroid Neural Network.

6. The system for processing interactive content of claim 3 , wherein the neural network comprises a convolutional neural network.

7. The system for processing interactive content of claim 6 , wherein the convolutional neural network (a) comprises at least three intermediate layers, and (b) performs operations that implement a Latent Dirichlet Allocation model.

8. The system for processing interactive content of claim 3 , wherein the neural network comprises a recurrent neural network having a long short-term memory neural network architecture.

9. The system for processing interactive content of claim 2 , wherein:

(a) the network computing device comprises a first neural network that is used to execute the subject classification analysis; and

(b) the network computing device comprises a second neural network that is used to execute the sentiment analysis.

10. The system for processing interactive content of claim 9 , wherein the first neural network performs operations that implement a Kmeans clustering analysis to execute the subject classification analysis.

11. The system for processing interactive content of claim 3 , wherein

(a) a labeling analysis is performed on a training set of content data files to generate annotated content data files;

(b) the network computing device processes the training set of content data files by performing a subject classification analysis that generates training subject classification identifiers;

(c) comparing the training subject classification identifiers against the annotated training set content data files to generate an error rate; and

(d) adjusting parameters of the neural network to reduce the error rate.

Assignments (2)

Continuity (1)

References Cited (23)