# US 10,872,303 B2 · Generating semantic representations for organizations and computing peer organization groups

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

- Number: US 10,872,303 B2
- Document ID: US10872303B2
- Status: Granted
- Kind: B2
- Application number: 15/994,148
- Filed: 2018-05-31
- Published: 2019-12-05
- Granted: 2020-12-22
- Earliest claimed: 2018-05-31
- Assignees: Microsoft Technology Licensing, LLC
- Applicants: Microsoft Technology Licensing, LLC
- Inventors: Yiqun Liu
- Web page: https://patentgrep.com/patent/US10872303B2

## CPC

- G06Q 10/067: Enterprise or organisation modelling
- H04L 63/0428: {wherein the data content is protected, e.g. by encrypting or encapsulating the payload}
- G06F 18/23213: with fixed number of clusters, e.g. K-means clustering
- H04L 63/0421: {Anonymous communication, i.e. the party's identifiers are hidden from the other party or parties, e.g. using an anonymizer}
- 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
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

In an example embodiment, a confidential data architecture is extended by providing components for reliably inferring confidential data insights (e.g., median salary) for cohorts with little or no actual submitted confidential data. This is performed by inferring confidential data values based on organizations that are considered to be peers to the organization of interest. This solution involves two parts: the generation of a novel, semantic representation (embedding) of organizations to be used to compute a similarity measure between any two organizations, and the use of the semantic representation to compute a peer organization group of a given company.

## Claims

### Claim 1

A system comprising:
a non-transitory computer-readable medium having instructions stored thereon, which, when executed by a processor, cause the system to:
obtain, via one or more graphical user interface, a plurality of electronic data submissions, each electronic data submission having user identification information and a corresponding confidential data value;
store each of the electronic data submissions in a row of a submission table by encrypting the user identification information using a first cryptographic key and storing the encrypted user identification information in a first column of the submission table, and encrypting the confidential data value using a second cryptographic key different than the first cryptographic key and storing the encrypted confidential data value in a second column of the submission table, thereby deidentifying the confidential data values;
obtain, for a first organization, information about transitions made by employees of the first organization to another organization or from another organization to the first organization;
calculate, for a particular organization of a plurality of organizations other than the first organization, a total number of transitions between the particular organization and the first organization;
for a combination of the first organization and the particular organization in the plurality of organizations other than the first organization, calculate a peer score based on the total number of transitions between the particular organization and the first organization;
identify the particular organization as part of a peer organization group with the first organization based on the calculated peer scores;
obtain a deidentified set of confidential data values for the first organization using the submission table;
determine that the deidentified set of confidential data values for the first organization contains fewer values than a predetermined threshold for providing a statistical insight;
in response to the determining, infer one or more confidential data values for the first organization using one or more confidential data values for one or more organizations, other than the first organization, in the peer organization group;
calculate a statistical insight based on the inferred one or more confidential data values; and
cause display of the statistical insight in a graphical user interface rendered on a display.

### Claim 2 (depends on 1)

The system of claim 1, wherein the calculating the peer score includes using negative sampling by randomly retrieving information about organizations not having any employees with transitions to or from the first organization.

### Claim 3 (depends on 1)

The system of claim 1, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using matrix factorization to generate organization group clusters.

### Claim 4 (depends on 1)

The system of claim 1, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using K-means clustering to generate organization group clusters.

### Claim 5 (depends on 1)

The system of claim 1; wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using Gaussian Process classification to generate organization group clusters.

### Claim 6 (depends on 1)

The system of claim 1, wherein the identifying comprises:
selecting a preset number of organizations having highest peer scores for the first organization for inclusion in the peer organization group.

### Claim 7 (depends on 1)

The system of claim 1, wherein the identifying comprises:
selecting organizations having peer scores for the first organization exceeding a predetermined threshold for inclusion in the peer organization group.

### Claim 8

A computerized method, comprising:
obtaining, via one or more graphical user interface, a plurality of electronic data submissions, each electronic data submission having user identification information and a corresponding confidential data value;
storing each of the electronic data submissions in a row of a submission table by encrypting the user identification information using a first cryptographic key and storing the encrypted user identification information in a first column of the submission table, and encrypting the confidential data value using a second cryptographic key different than the first cryptographic key and storing the encrypted confidential data value in a second column of the submission table, thereby deidentifying the confidential data values;
obtaining, for a first organization, information about transitions made by employees of the first organization to another organization or from another organization to the first organization;
calculating, for a particular organization of a plurality of organizations other than the first organization, a total number of transitions between the particular organization and the first organization;
for a combination of the first organization and the particular organization in the plurality of organizations other than the first organization, calculating a peer score based on the total number of transitions between the particular organization and the first organization;
identifying the particular organization as part of a peer organization group with the first organization based on the calculated peer scores;
obtaining a deidentified set of confidential data values for the first organization using the submission table;
determining that the deidentified set of confidential data values for the first organization contains fewer values than a predetermined threshold for providing a statistical insight;
in response to the determining, inferring one or more confidential data values for the first organization using one or more confidential data values for one or more organizations, other than the first organization, in the peer organization group;
calculating a statistical insight based on the inferred one or more confidential data values; and
causing display of the statistical insight in a graphical user interface rendered on a display.

### Claim 9 (depends on 8)

The method of claim 8, wherein the calculating the peer score includes using negative sampling by randomly retrieving information about organizations not having any employees with transitions to or from the first organization.

### Claim 10 (depends on 8)

The method of claim 8, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using matrix factorization to generate organization group clusters.

### Claim 11 (depends on 8)

The method of claim 8, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using K-means clustering to generate organization group clusters.

### Claim 12 (depends on 8)

The method of claim 8, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using Gaussian Process classification to generate organization group clusters.

### Claim 13 (depends on 8)

The method of claim 8, wherein the identifying comprises:
selecting a preset number of organizations having highest peer scores for the first organization for inclusion in the peer organization group.

### Claim 14 (depends on 8)

The method of claim 8, wherein the identifying comprises:
selecting organizations having peer scores for the first organization exceeding a predetermined threshold for inclusion in the peer organization group.

### Claim 15

A non-transitory machine-readable storage medium comprising instructions, which when implemented by one or more machines, cause the one or more machines to perform operations comprising:
obtaining, via one or more graphical user interface, a plurality of electronic data submissions, each electronic data submission having user identification information and a corresponding confidential data value;
storing each of the electronic data submissions in a row of a submission table by encrypting the user identification information using a first cryptographic key and storing the encrypted user identification information in a first column of the submission table, and encrypting the confidential data value using a second cryptographic key different than the first cryptographic key and storing the encrypted confidential data value in a second column of the submission table, thereby deidentifying the confidential data values;
obtaining, for a first organization, information about transitions made by employees of the first organization to another organization or from another organization to the first organization;
calculating, for a particular organization of a plurality of organizations other than the first organization, a total number of transitions between the particular organization and the first organization;
for a combination of the first organization and the particular organization in the plurality of organizations other than the first organization, calculating a peer score based on the total number of transitions between the particular organization and the first organization;
identifying the particular organization as part of a peer organization group with the first organization based on the calculated peer scores;
obtaining a deidentified set of confidential data values for the first organization using the submission table;
determining that the deidentified set of confidential data values for the first organization contains fewer values than a predetermined threshold for providing a statistical insight;
in response to the determining, inferring one or more confidential data values for the first organization using one or more confidential data values for one or more organizations, other than the first organization, in the peer organization group;
calculating a statistical insight based on the inferred one or more confidential data values; and
causing display of the statistical insight in a graphical user interface rendered on a display.

### Claim 16 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the calculating the peer score includes using negative sampling by randomly retrieving information about organizations not having any employees with transitions to or from the first organization.

### Claim 17 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using matrix factorization to generate organization group dusters.

### Claim 18 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using K-means clustering to generate organization group clusters.

### Claim 19 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the identifying comprises:
generating a matrix of organization-organization peer scores; and
using Gaussian Process classification to generate organization group clusters.

### Claim 20 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the identifying comprises:
selecting a preset of number organizations having highest peer scores for the first organization for inclusion in the peer organization group.