# US 10,482,185 B1 · Methods and arrangements to adjust communications

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 10,482,185 B1
- Document ID: US10482185B1
- Status: Granted
- Kind: B1
- Application number: 16/286,999
- Filed: 2019-02-27
- Granted: 2019-11-19
- Earliest claimed: 2019-02-27
- Assignees: Capital One Services, LLC
- Applicants: Capital One Services, LLC
- Inventors: Fardin Abdi Taghi Abad; Austin Grant Walters; Jeremy Edward Goodsitt; Reza Farivar; Vincent Pham; Anh Truong
- Web page: https://patentgrep.com/patent/US10482185B1

## CPC

- G06N 20/00: Machine learning
- G06F 16/316: {Indexing structures}
- G06F 40/247: Thesauruses; Synonyms
- G06F 40/268: Morphological analysis
- G06F 40/30: Semantic analysis
- G06F 40/47: Machine-assisted translation, e.g. using translation memory
- G06F 40/56: Natural language generation
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/063: using electronic means
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- G06Q 30/016: After-sales
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR

## Abstract

Logic may adjust communications between customers. Logic may cluster customers into a first group associated with a first subset of synonyms and a second group associated with a second subset of the synonyms. Logic may associate a first tag with the first group and with each of the synonyms of the first subset. Logic may associate a second tag with the second group and with each of the synonyms of the second subset. Logic may associate one or more models with pairs of the groups. A first pair may comprise the first group and the second group. The first model associated with the first pair may adjust words in communications between the first group and the second group, based on the synonyms associated with the first pair, by replacement of words in a communication between customers of the first subset and customers of the second sub set.

## Claims

### Claim 1

An apparatus comprising:
memory; and
logic circuitry coupled with the memory, the logic circuitry configured to:
cluster customers into at least two groups based on customer data, wherein the at least two groups comprise a first group associated with a first subset of synonyms and a second group associated with a second subset of the synonyms, wherein each customer is associated with one of the groups;
associate a first unique tag with the first group and with each of the synonyms of the first subset;
associate a second unique tag with the second group and with each of the synonyms of the second subset; wherein the first subset and the second subset comprise unique subsets;
associate one or more models with pairs of the groups, wherein a first pair of the pairs comprises the first group and the second group, and wherein a first model of the models is associated with the first pair; and
replace, by the first model, at least a first synonym of the first subset with a second synonym of the second subset in a first communication from a first customer of the first group to a second customer of the second group.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein each of the synonyms is associated with one or more of the groups.

### Claim 3 (depends on 2)

The apparatus of claim 2, wherein a cluster model comprises a statistical model or a machine learning model to associate each of the customers with one of the groups based on the customer data.

### Claim 4 (depends on 3)

The apparatus of claim 3, wherein the customer data comprises financial information about the customers, geographical information about the customers, historical information about the customers, communications prepared by the customers, social media associated with the customers, other information associated with the customers, or a combination thereof.

### Claim 5 (depends on 1)

The apparatus of claim 1, wherein communications comprise writings authored by the customers, wherein the writings comprise one or more of the synonyms.

### Claim 6 (depends on 1)

The apparatus of claim 1, wherein a cluster model associates the first customer with the first group based on the first subset of the synonyms and associates the first subset of synonyms with the first group based on synonyms common to each of the customers associated with the first group.

### Claim 7 (depends on 1)

The apparatus of claim 1, wherein the first model selects the first synonym to replace the second synonym based on training with the customer data, wherein the customer data comprises the subsets of synonyms.

### Claim 8 (depends on 1)

The apparatus of claim 1, wherein each of the groups comprises a node in a graph database and each of the models is associated with an edge of one of the nodes in the graph database.

### Claim 9 (depends on 1)

The apparatus of claim 1, further comprising a monitor model configured to track error rates associated with replacement of synonyms in communications between the customers, wherein the error rates represent misunderstandings by the customers in response to the replacement of synonyms, the monitor model configured to cause changes in associations of one or more of the customers with the groups based on one or more of the error rates exceeding a threshold.

### Claim 10

A non-transitory storage medium containing instructions, which when executed by a processor, cause the processor to:
form a first group with a first subset of customers and a second group with a second subset of the customers; wherein the first subset of the customers and the second subset of the customers are mutually exclusive;
associate a first unique tag with the first group and with a first subset of synonyms;
associate a second unique tag with the second group and with a second subset of the synonyms;
train at least one model to replace synonyms in communications between the first group and the second group based on customer data associated with the first group and the second group; and
replace, by the at least one model, at least a first synonym from the first subset of synonyms with a second synonym from the second subset of synonyms in a first communication from a first customer of the first subset of customers to a second customer of the second subset of customers.

### Claim 11 (depends on 10)

The non-transitory storage medium of claim 10, the processor further being caused to cluster the customers into the groups based on the customer data associated with the first group and the second group.

### Claim 12 (depends on 11)

The non-transitory storage medium of claim 11, the processor further being caused to to associate each of the customers with one of the groups based on tags associated with each of the customers.

### Claim 13 (depends on 11)

The non-transitory storage medium of claim 11, wherein the customer data comprises financial information, geographical information, historical information, communications, social media, or a combination thereof.

### Claim 14 (depends on 13)

The non-transitory storage medium of claim 13, wherein communications comprise writings authored by the customers, and wherein the writings comprise one or more of the synonyms.

### Claim 15 (depends on 10)

The non-transitory storage medium of claim 10, the processor further being caused to associate the first customer with the first group based on the first subset of the synonyms and associate the first subset of synonyms with the first group based on synonyms common to each of the customers associated with the first group.

### Claim 16 (depends on 10)

The non-transitory storage medium of claim 10, the processor further being caused to select the first synonym to replace the second synonym based on training with the customer data, wherein the customer data comprises the first subset of synonyms and second subset of synonyms.

### Claim 17 (depends on 10)

The non-transitory storage medium of claim 10, wherein the first group comprises a first node in a graph database and the at least one model is associated with an edge of the first node in the graph database.

### Claim 18 (depends on 17)

The non-transitory storage medium of claim 17, the processor further being caused to track error rates associated with replacement of synonyms in communications between the customers, wherein the error rates represent misunderstandings by the customers in response to the replacement of synonyms in communications, and the processor further being caused to change associations of one or more of the customers with one of the groups based on one or more of the error rates exceeding a threshold.

### Claim 19 (depends on 18)

The non-transitory storage medium of claim 18, the processor further being caused to establish a new node after one or more changes in associations between one or more of the customers and one of or both of the first group and the second group.

### Claim 20

A system comprising:
memory; and
logic circuitry coupled with the memory, the logic circuitry configured to:
provide a database including a set of synonyms, each of the synonyms associated with one or more tags, each of the tags associated with one or more nodes of the database;
receive, via an application programming interface (API), an instruction to replace one or more words in at least one of a document drafted in a text editor, an identification of an author of the document, an identification of a recipient of the document, or any combination thereof;
identify a first tag associated with the identification of the recipient and a second tag associated with the identification of the author;
select a model based on an association between the first tag, the second tag, and the model;
process, via the API, text of the document to identify one or more words associated with the second tag;
identify, by the model, synonyms associated with the first tag to replace the one or more words associated with the second tag; and
replace, via the API, the one or more words associated with the second tag with the synonyms identified by the model.