# US 2022/0284520 A1 · System and method for recommending matches using machine learning

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

- Number: US 2022/0284520 A1
- Document ID: US20220284520A1
- Status: Published application
- Kind: A1
- Application number: 17/592,454
- Filed: 2022-02-03
- Published: 2022-09-08
- Earliest claimed: 2021-03-04
- Inventors: Timothy Dirk Stevens
- Web page: https://patentgrep.com/patent/US20220284520A1

## CPC

- G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06N 3/04: Architecture, e.g. interconnection topology
- 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 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/08: Learning methods
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn
- G06Q 10/42: Determination of affinities or common interests between users
- G06N 3/045: Combinations of networks
- G06N 3/048: Activation functions
- G06N 20/20: Ensemble learning
- G06Q 10/105: Human resources
- G06Q 30/0203: Market surveys; Market polls
- G06V 40/168: {Feature extraction; Face representation}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- 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
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A system and method for recommending matches between persons are provided. Data processing is performed using artificial intelligence technology. A supervised machine learning engine is trained from empirical data about existing relationships which have been evaluated as to quality of the relationships. Quality of candidate relationships is calculated as an output of the supervised machine learning engine when provided input data of attributes of two candidate persons. Likelihood of a successful relationship between two candidate persons is predicted by comparing the calculated quality of the candidate relationship against a threshold. The prediction in this learning task may be made by a neural network. A user is notified of a candidate match that is likely to become a successful relationship.

## Claims

### Claim 1

A method for recommending persons for a dating interpersonal relationship, comprising:
training a supervised machine learning engine from a database of multiple existing relationships;
wherein the database of data describing multiple existing relationships comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the database of data describing multiple existing relationships comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing additional attributes provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
inputting data into the trained supervised machine learning engine attributes of a first person and attributes of a plurality of candidate persons for the relationship;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of a potential relationship between the first person and each candidate person in the plurality of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining a likelihood of a successful relationship for each candidate person by comparing the numerical evaluation against a threshold; and
recommending at least one candidate person of the plurality of candidate persons to the first person based on results of the comparing.

### Claim 2 (depends on 1)

The method of claim 1, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 3 (depends on 1)

The method of claim 1, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 4 (depends on 1)

The method of claim 1, wherein the additional attributes processed by the second neural network further comprise respective facial images of the first person and the second person.

### Claim 5 (depends on 1)

The method of claim 1, wherein the additional attributes processed by the second neural network further comprise respective videos of the first person and the second person.

### Claim 6 (depends on 1)

The method of claim 1, wherein the additional attributes processed by the second neural network further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 7 (depends on 2)

The method of claim 2, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 8 (depends on 4)

The method of claim 4, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective videos of the first person and the second person.

### Claim 9 (depends on 5)

The method of claim 5, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 10 (depends on 6)

The method of claim 6, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective facial images of the first person and the second person.

### Claim 11 (depends on 7)

The method of claim 7, wherein the additional existing relationships include relationships formed based on previous recommendations.

### Claim 12 (depends on 8)

The method of claim 8, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective videos of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 13 (depends on 9)

The method of claim 9, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 14 (depends on 10)

The method of claim 10, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective facial images of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 15 (depends on 12)

The method of claim 12, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 16 (depends on 13)

The method of claim 13, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 17 (depends on 14)

The method of claim 14, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 18 (depends on 15)

The method of claim 15, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 19 (depends on 18)

The method of claim 18, wherein the supervised machine learning engine further comprises a fourth neural network, different from the first and second and third neural networks, the fourth neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the fourth neural network has multiple hidden layers.

### Claim 20 (depends on 19)

The method of claim 19, wherein the prediction is further based on outputs of the first and second and third and fourth neural networks.

### Claim 21

A method for recommending persons for an intimate interpersonal relationship, comprising:
training a supervised machine learning engine from a database of multiple existing relationships;
wherein the database of data describing multiple existing relationships comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the database of data describing multiple existing relationships comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing additional attributes provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
inputting data into the trained supervised machine learning engine attributes of a first person and attributes of a plurality of candidate persons for the relationship;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of a potential relationship between the first person and each candidate person in the plurality of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining a likelihood of a successful relationship for each candidate person by comparing the numerical evaluation against a threshold; and
recommending at least one candidate person of the plurality of candidate persons to the first person based on results of the comparing.

### Claim 22 (depends on 21)

The method of claim 21, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 23 (depends on 21)

The method of claim 21, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 24 (depends on 21)

The method of claim 21, wherein the additional attributes processed by the second neural network further comprise respective facial images of the first person and the second person.

### Claim 25 (depends on 21)

The method of claim 21, wherein the additional attributes processed by the second neural network further comprise respective videos of the first person and the second person.

### Claim 26 (depends on 21)

The method of claim 21, wherein the additional attributes processed by the second neural network further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 27 (depends on 22)

The method of claim 22, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 28 (depends on 24)

The method of claim 24, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective videos of the first person and the second person.

### Claim 29 (depends on 25)

The method of claim 25, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 30 (depends on 26)

The method of claim 26, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective facial images of the first person and the second person.

### Claim 31 (depends on 27)

The method of claim 27, wherein the additional existing relationships include relationships formed based on previous recommendations.

### Claim 32 (depends on 28)

The method of claim 28, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective videos of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 33 (depends on 29)

The method of claim 29, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 34 (depends on 30)

The method of claim 30, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective facial images of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 35 (depends on 32)

The method of claim 32, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 36 (depends on 33)

The method of claim 33, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 37 (depends on 34)

The method of claim 34, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 38 (depends on 35)

The method of claim 35, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 39 (depends on 38)

The method of claim 38, wherein the supervised machine learning engine further comprises a fourth neural network, different from the first and second and third neural networks, the fourth neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the fourth neural network has multiple hidden layers.

### Claim 40 (depends on 39)

The method of claim 39, wherein the prediction is further based on outputs of the first and second and third and fourth neural networks.

### Claim 41

A system for recommending persons for a dating interpersonal relationship, the system comprising:
at least one database storing data for multiple existing relationships and attributes of persons who are candidate persons for the relationship;
wherein the at least one database comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the at least one database comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
a user interface;
at least one processor coupled to the at least one database and user interface and configured for:
training a supervised machine learning engine from the at least one database;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing additional attributes provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
retrieving from the at least one database attributes of a first person seeking a relationship;
querying the at least one database to retrieve attributes of each candidate person for the relationship with the first person;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of each potential relationship between the first person seeking a relationship and each candidate person in the query results of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining the likelihood of successful relationships by comparing the numerical evaluations against a threshold; and
recommending to the first person seeking a relationship at least one candidate person based on results of the comparing.

### Claim 42 (depends on 41)

The system of claim 41, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 43 (depends on 41)

The system of claim 41, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 44 (depends on 41)

The system of claim 41, wherein the additional attributes processed by the second neural network further comprise respective facial images of the first person and the second person.

### Claim 45 (depends on 41)

The system of claim 41, wherein the additional attributes processed by the second neural network further comprise respective videos of the first person and the second person.

### Claim 46 (depends on 41)

The system of claim 41, wherein the additional attributes processed by the second neural network further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 47 (depends on 41)

The system of claim 41, wherein querying the at least one database selects candidate persons who satisfy matching criteria provided by the first person seeking a relationship.

### Claim 48 (depends on 41)

The system of claim 41, wherein querying the at least one database selects candidate persons whose matching criteria are satisfied by the first person seeking a relationship.

### Claim 49 (depends on 41)

The system of claim 41, wherein querying the at least one database selects candidate persons who satisfy matching criteria provided by the first person seeking a relationship logically ANDed with candidates whose matching criteria are satisfied by the first person seeking a relationship.

### Claim 50 (depends on 41)

The system of claim 41, wherein querying the at least one database selects candidate persons who satisfy matching criteria provided by the first person seeking a relationship logically ORed with candidates whose matching criteria are satisfied by the first person seeking a relationship.

### Claim 51 (depends on 41)

The system of claim 41, wherein being considered as a candidate when querying the at least one database includes persons who do not meet matching criteria but are within a threshold of meeting the matching criteria.

### Claim 52 (depends on 42)

The system of claim 42, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 53 (depends on 44)

The system of claim 44, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective videos of the first person and the second person.

### Claim 54 (depends on 45)

The system of claim 45, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 55 (depends on 46)

The system of claim 46, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective facial images of the first person and the second person.

### Claim 56 (depends on 52)

The system of claim 52, wherein the additional existing relationships include relationships formed based on previous recommendations.

### Claim 57 (depends on 53)

The system of claim 53, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective videos of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 58 (depends on 54)

The system of claim 54, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 59 (depends on 55)

The system of claim 55, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective facial images of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 60 (depends on 57)

The system of claim 57, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 61 (depends on 58)

The system of claim 58, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 62 (depends on 59)

The system of claim 59, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 63 (depends on 60)

The system of claim 60, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 64 (depends on 63)

The system of claim 63, wherein the supervised machine learning engine further comprises a fourth neural network, different from the first and second and third neural networks, the fourth neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the fourth neural network has multiple hidden layers.

### Claim 65 (depends on 64)

The system of claim 64, wherein the prediction is further based on outputs of the first and second and third and fourth neural networks.

### Claim 66

A system for recommending persons for an intimate interpersonal relationship, the system comprising:
at least one database storing data for multiple existing relationships and attributes of persons who are candidate persons for the relationship;
wherein the at least one database comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the at least one database comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
a user interface;
at least one processor coupled to the at least one database and user interface and configured for:
training a supervised machine learning engine from the at least one database;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing additional attributes provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
retrieving from the at least one database attributes of a first person seeking a relationship;
querying the at least one database to retrieve attributes of each candidate person for the relationship with the first person;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of each potential relationship between the first person seeking a relationship and each candidate person in the query results of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining the likelihood of successful relationships by comparing the numerical evaluations against a threshold; and
recommending to the first person seeking a relationship at least one candidate person based on results of the comparing.

### Claim 67 (depends on 66)

The system of claim 66, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 68 (depends on 66)

The system of claim 66, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 69 (depends on 66)

The system of claim 66, wherein the additional attributes processed by the second neural network further comprise respective facial images of the first person and the second person.

### Claim 70 (depends on 66)

The system of claim 66, wherein the additional attributes processed by the second neural network further comprise respective videos of the first person and the second person.

### Claim 71 (depends on 66)

The system of claim 66, wherein the additional attributes processed by the second neural network further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 72 (depends on 66)

The system of claim 66, wherein querying the at least one database selects candidate persons who satisfy matching criteria provided by the first person seeking a relationship.

### Claim 73 (depends on 66)

The system of claim 66, wherein querying the at least one database selects candidate persons whose matching criteria are satisfied by the first person seeking a relationship.

### Claim 74 (depends on 66)

The system of claim 66, wherein querying the at least one database selects candidate persons who satisfy matching criteria provided by the first person seeking a relationship logically ANDed with candidates whose matching criteria are satisfied by the first person seeking a relationship.

### Claim 75 (depends on 66)

The system of claim 66, wherein querying the at least one database selects candidate persons who satisfy matching criteria provided by the first person seeking a relationship logically ORed with candidates whose matching criteria are satisfied by the first person seeking a relationship.

### Claim 76 (depends on 66)

The system of claim 66, wherein being considered as a candidate when querying the at least one database includes persons who do not meet matching criteria but are within a threshold of meeting the matching criteria.

### Claim 77 (depends on 67)

The system of claim 67, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 78 (depends on 69)

The system of claim 69, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective videos of the first person and the second person.

### Claim 79 (depends on 70)

The system of claim 70, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 80 (depends on 71)

The system of claim 71, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective facial images of the first person and the second person.

### Claim 81 (depends on 77)

The system of claim 77, wherein the additional existing relationships include relationships formed based on previous recommendations.

### Claim 82 (depends on 78)

The system of claim 78, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective videos of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 83 (depends on 79)

The system of claim 79, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 84 (depends on 80)

The system of claim 80, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective facial images of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 85 (depends on 82)

The system of claim 82, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 86 (depends on 83)

The system of claim 83, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 87 (depends on 84)

The system of claim 84, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 88 (depends on 85)

The system of claim 85, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise data mined from respective social networking accounts of the first person and the second person.

### Claim 89 (depends on 88)

The system of claim 88, wherein the supervised machine learning engine further comprises a fourth neural network, different from the first and second and third neural networks, the fourth neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the fourth neural network has multiple hidden layers.

### Claim 90 (depends on 89)

The system of claim 89, wherein the prediction is further based on outputs of the first and second and third and fourth neural networks.

### Claim 91

A method for recommending persons for an advisor relationship with a client, comprising:
training a supervised machine learning engine from a database of multiple existing relationships;
wherein the database of data describing multiple existing relationships comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the database of data describing multiple existing relationships comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective videos of the first person and the second person;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing the videos provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
inputting data into the trained supervised machine learning engine attributes of a first person and attributes of a plurality of candidate persons for the relationship;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of a potential relationship between the first person and each candidate person in the plurality of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining a likelihood of a successful relationship for each candidate person by comparing the numerical evaluation against a threshold; and
recommending at least one candidate person of the plurality of candidate persons to the first person based on results of the comparing.

### Claim 92 (depends on 91)

The method of claim 91, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 93 (depends on 91)

The method of claim 91, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 94 (depends on 92)

The method of claim 92, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 95

A method for recommending persons for a teacher relationship with a student, comprising:
training a supervised machine learning engine from a database of multiple existing relationships;
wherein the database of data describing multiple existing relationships comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the database of data describing multiple existing relationships comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective videos of the first person and the second person;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing the videos provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
inputting data into the trained supervised machine learning engine attributes of a first person and attributes of a plurality of candidate persons for the relationship;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of a potential relationship between the first person and each candidate person in the plurality of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining a likelihood of a successful relationship for each candidate person by comparing the numerical evaluation against a threshold; and
recommending at least one candidate person of the plurality of candidate persons to the first person based on results of the comparing.

### Claim 96 (depends on 95)

The method of claim 95, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 97 (depends on 95)

The method of claim 95, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 98 (depends on 96)

The method of claim 96, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 99

A method for recommending persons for a relationship between an employer and an employee, comprising:
training a supervised machine learning engine from a database of multiple existing relationships;
wherein the database of data describing multiple existing relationships comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the database of data describing multiple existing relationships comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing additional attributes provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
inputting data into the trained supervised machine learning engine attributes of a first person and attributes of a plurality of candidate persons for the relationship;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of a potential relationship between the first person and each candidate person in the plurality of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining a likelihood of a successful relationship for each candidate person by comparing the numerical evaluation against a threshold; and
recommending at least one candidate person of the plurality of candidate persons to the first person based on results of the comparing.

### Claim 100 (depends on 99)

The method of claim 99, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 101 (depends on 99)

The method of claim 99, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 102 (depends on 99)

The method of claim 99, wherein the additional attributes processed by the second neural network further comprise respective facial images of the first person and the second person.

### Claim 103 (depends on 99)

The method of claim 99, wherein the additional attributes processed by the second neural network further comprise respective videos of the first person and the second person.

### Claim 104 (depends on 100)

The method of claim 100, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 105 (depends on 102)

The method of claim 102, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective videos of the first person and the second person.

### Claim 106 (depends on 105)

The method of claim 105, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective videos of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 107 (depends on 106)

The method of claim 106, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 108

A method for recommending persons for a player relationship in at least one multiplayer game, comprising:
training a supervised machine learning engine from a database of multiple existing relationships;
wherein the database of data describing multiple existing relationships comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the database of data describing multiple existing relationships comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise data mined from respective social networking accounts of the first person and the second person;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
inputting data into the trained supervised machine learning engine attributes of a first person and attributes of a plurality of candidate persons for the relationship;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of a potential relationship between the first person and each candidate person in the plurality of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining a likelihood of a successful relationship for each candidate person by comparing the numerical evaluation against a threshold; and
recommending at least one candidate person of the plurality of candidate persons to the first person based on results of the comparing.

### Claim 109 (depends on 108)

The method of claim 108, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 110 (depends on 108)

The method of claim 108, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 111 (depends on 109)

The method of claim 109, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 112

A system for recommending persons for an advisor relationship with a client, the system comprising:
at least one database storing data for multiple existing relationships and attributes of persons who are candidate persons for the relationship;
wherein the at least one database comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the at least one database comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective videos of the first person and the second person;
a user interface;
at least one processor coupled to the at least one database and user interface and configured for:
training a supervised machine learning engine from the at least one database;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing respective videos provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
retrieving from the at least one database attributes of a first person seeking a relationship;
querying the at least one database to retrieve attributes of each candidate person for the relationship with the first person;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of each potential relationship between the first person seeking a relationship and each candidate person in the query results of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining the likelihood of successful relationships by comparing the numerical evaluations against a threshold; and
recommending to the first person seeking a relationship at least one candidate person based on results of the comparing.

### Claim 113 (depends on 112)

The system of claim 112, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 114 (depends on 112)

The system of claim 112, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 115 (depends on 113)

The system of claim 113, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 116

A system for recommending persons for a teacher relationship with a student, the system comprising:
at least one database storing data for multiple existing relationships and attributes of persons who are candidate persons for the relationship;
wherein the at least one database comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the at least one database comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective videos of the first person and the second person;
a user interface;
at least one processor coupled to the at least one database and user interface and configured for:
training a supervised machine learning engine from the at least one database;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing respective videos provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
retrieving from the at least one database attributes of a first person seeking a relationship;
querying the at least one database to retrieve attributes of each candidate person for the relationship with the first person;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of each potential relationship between the first person seeking a relationship and each candidate person in the query results of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining the likelihood of successful relationships by comparing the numerical evaluations against a threshold; and
recommending to the first person seeking a relationship at least one candidate person based on results of the comparing.

### Claim 117 (depends on 116)

The system of claim 116, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 118 (depends on 116)

The system of claim 116, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 119 (depends on 117)

The system of claim 117, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 120

A system for recommending persons for a relationship between an employer and an employee, the system comprising:
at least one database storing data for multiple existing relationships and attributes of persons who are candidate persons for the relationship;
wherein the at least one database comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the at least one database comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
a user interface;
at least one processor coupled to the at least one database and user interface and configured for:
training a supervised machine learning engine from the at least one database;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing additional attributes provided by the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
retrieving from the at least one database attributes of a first person seeking a relationship;
querying the at least one database to retrieve attributes of each candidate person for the relationship with the first person;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of each potential relationship between the first person seeking a relationship and each candidate person in the query results of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining the likelihood of successful relationships by comparing the numerical evaluations against a threshold; and
recommending to the first person seeking a relationship at least one candidate person based on results of the comparing.

### Claim 121 (depends on 120)

The system of claim 120, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 122 (depends on 120)

The system of claim 120, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 123 (depends on 120)

The system of claim 120, wherein the additional attributes processed by the second neural network further comprise respective facial images of the first person and the second person.

### Claim 124 (depends on 120)

The system of claim 120, wherein the additional attributes processed by the second neural network further comprise respective videos of the first person and the second person.

### Claim 125 (depends on 121)

The system of claim 121, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.

### Claim 126 (depends on 123)

The system of claim 123, wherein the attributes of the first person in the relationship and the second person in the relationship further comprise respective videos of the first person and the second person.

### Claim 127 (depends on 126)

The system of claim 126, wherein the supervised machine learning engine further comprises a third neural network, different from the first and second neural networks, the third neural network processing the respective videos of the first person and the second person in each relationship, where the third neural network has multiple hidden layers.

### Claim 128 (depends on 127)

The system of claim 127, wherein the prediction is further based on outputs of the first and second and third neural networks.

### Claim 129

A system for recommending persons for a player relationship in at least one multiplayer game, the system comprising:
at least one database storing data for multiple existing relationships and attributes of persons who are candidate persons for the relationship;
wherein the at least one database comprises successful relationships and unsuccessful relationships;
wherein the data describing each relationship in the at least one database comprises attributes of a first person in the relationship, attributes of a second person in the relationship and an evaluation of the relationship;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise respective answers provided by the first person and the second person to survey questions;
wherein the attributes of the first person in the relationship and the second person in the relationship comprise data mined from respective social networking accounts of the first person and the second person;
a user interface;
at least one processor coupled to the at least one database and user interface and configured for:
training a supervised machine learning engine from the at least one database;
wherein the supervised machine learning engine comprises a first neural network, the first neural network processing the answers to survey questions provided by the first person and the second person in each relationship, where the first neural network has multiple hidden layers;
wherein the supervised machine learning engine comprises a second neural network, different from the first neural network, the second neural network processing the data mined from respective social networking accounts of the first person and the second person in each relationship, where the second neural network has multiple hidden layers;
retrieving from the at least one database attributes of a first person seeking a relationship;
retrieving from the at least one database attributes of a first person seeking a relationship;
querying the at least one database to retrieve attributes of each candidate person for the relationship with the first person;
utilizing the trained supervised machine learning engine to predict a numerical evaluation of each potential relationship between the first person seeking a relationship and each candidate person in the query results of candidate persons;
wherein the prediction is based on outputs of the first and second neural networks;
determining the likelihood of successful relationships by comparing the numerical evaluations against a threshold; and
recommending to the first person seeking a relationship at least one candidate person based on results of the comparing.

### Claim 130 (depends on 129)

The system of claim 129, wherein training the supervised machine learning engine comprises minimizing the difference between the numerical evaluation of a relationship output by the supervised machine learning engine and the numerical evaluation of that relationship obtained from the database of data describing multiple existing relationships.

### Claim 131 (depends on 129)

The system of claim 129, wherein the neural networks comprise a heterogeneous hybrid neural network.

### Claim 132 (depends on 130)

The system of claim 130, further comprising updating the training of the supervised machine learning engine using data of additional existing relationships.
