# US 2021/0334458 A1 · Systems for Role Classification

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

- Number: US 2021/0334458 A1
- Document ID: US20210334458A1
- Status: Published application
- Kind: A1
- Application number: 16/859,780
- Filed: 2020-04-27
- Published: 2021-10-28
- Earliest claimed: 2020-04-27
- Assignees: Adobe Inc.
- Applicants: Adobe Inc.
- Inventors: Ajay Jain; Sanjeev Tagra; Sachin Soni; Niranjan Shivanand Kumbi; Eric Andrew Kienle; Ajay Awatramani; Abhishek Jain
- Web page: https://patentgrep.com/patent/US20210334458A1

## CPC

- G06F 40/216: using statistical methods
- G06F 40/279: Recognition of textual entities
- G06F 40/284: Lexical analysis, e.g. tokenisation or collocates
- G06F 40/30: Semantic analysis
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/00: Machine learning
- G06N 3/048: Activation functions
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

In implementations of systems for role classification, a computing device implements a role system to receive data describing a corpus of text that is associated with a user ID. Feature values of features are generated by a first machine learning model by processing the corpus of text, the features representing questions with respect to the corpus of text and the feature values representing answers to the questions included in the corpus of text. A classification of a role is generated by a second machine learning model by processing the feature values, the classification of the role indicating a relationship of the user ID with respect to a product or service. The role system outputs an indication of the classification of the role for display in a user interface of a display device.

## Claims

### Claim 1

In a digital medium environment for classifying a role, a method implemented by a computing device, the method comprising:
receiving, by the computing device, data describing a corpus of text that is associated with a user ID;
generating, by the computing device, feature values of features by a first machine learning model by processing the corpus of text, the features representing questions with respect to the corpus of text and the feature values representing answers to the questions included in the corpus of text;
generating, by the computing device, a classification of the role by a second machine learning model by processing the feature values, the classification of the role indicating a relationship of the user ID with respect to a product or service; and
outputting, by the computing device, an indication of the classification of the role for display in a user interface of a display device.

### Claim 2 (depends on 1)

The method as described in claim 1, further comprising receiving a user input defining the user ID, the user input received via the user interface.

### Claim 3 (depends on 1)

The method as described in claim 1, wherein the first machine learning model includes a long short term memory model.

### Claim 4 (depends on 1)

The method as described in claim 1, wherein the second machine learning model includes a logistic regression model.

### Claim 5 (depends on 1)

The method as described in claim 1, wherein the classification of the role is at least one of a decision maker, a champion, a detractor, or an influencer.

### Claim 6 (depends on 1)

The method as described in claim 1, wherein the indication includes a one-versus-rest probability.

### Claim 7 (depends on 1)

The method as described in claim 1, further comprising forming a communication for receipt by a client device associated with the user ID, the communication including digital content based on the classification of the role.

### Claim 8 (depends on 1)

The method as described in claim 1, wherein the features include a sentiment of text included in the corpus of text and a tone of text included in the corpus of text.

### Claim 9

In a digital medium environment to train a first machine learning model for role classification, a system comprising:
at least one computing device; and
a role module implemented at least partially in hardware of the at least one computing device to:
receive a first set of training data including a training classification and training feature values of features, the training features values generated by a second machine learning model by processing a corpus of text associated with an ID and the training classification is indicative of a relationship of the ID with respect to a product or service;
generate a candidate classification by processing the training feature values using the first machine learning model; and
train the first machine learning model by comparing the candidate classification to the training classification using a first loss function.

### Claim 10 (depends on 9)

The system as described in claim 9, wherein the role module is further implemented to:
receive a second set of training data including the corpus of text, the features, and ground truth feature values;
generate candidate feature values by processing the corpus of text and the features using the second machine learning model; and
train the second machine learning model by comparing the candidate feature values to the ground truth feature values using a second loss function.

### Claim 11 (depends on 9)

The system as described in claim 9, wherein the candidate classification is at least one of a decision maker, a champion, a detractor, or an influencer.

### Claim 12 (depends on 9)

The system as described in claim 9, wherein the ID is a client device ID.

### Claim 13 (depends on 9)

The system as described in claim 9, wherein the first machine learning model includes a logistic regression model and the second machine learning model includes a long short term memory model.

### Claim 14

A system comprising:
a compilation module implemented at least partially in hardware of a computing device to:
receive a user input defining a user ID for role classification; and
compile a corpus of text associated with the user ID;
a feature module implemented at least partially in the hardware of the computing device to generate feature values of features by processing the corpus of text, the features representing questions with respect to the corpus of text and the feature values representing answers to the questions included in the corpus of text;
a classification module implemented at least partially in the hardware of the computing device to generate a classification of a role by processing the feature values, the classification of the role indicating a relationship of the user ID with respect to a product or service; and
a display module implemented at least partially in the hardware of the computing device to output an indication of the classification of the role for display in a user interface.

### Claim 15 (depends on 14)

The system as described in claim 14, wherein the classification of the role is at least one of a decision maker, a champion, a detractor, or an influencer.

### Claim 16 (depends on 14)

The system as described in claim 14, wherein the indication includes a one-versus-rest probability.

### Claim 17 (depends on 14)

The system as described in claim 14, wherein the feature module includes a long short term memory model.

### Claim 18 (depends on 14)

The system as described in claim 14, wherein the classification module includes a logistic regression model.

### Claim 19 (depends on 14)

The system as described in claim 14, further comprising a communication module implemented at least partially in the hardware of the computing device to form a communication for receipt by a client device associated with the user ID, the communication including digital content based on the classification of the role.

### Claim 20 (depends on 14)

The system as described in claim 14, wherein the features include a sentiment of text included in the corpus of text and a tone of text included in the corpus of text.
