# US 2021/0064684 A1 · Search optimization based on relevant-parameter selection

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

- Number: US 2021/0064684 A1
- Document ID: US20210064684A1
- Status: Published application
- Kind: A1
- Application number: 16/553,375
- Filed: 2019-08-28
- Published: 2021-03-04
- Earliest claimed: 2019-08-28
- Assignees: Microsoft Technology Licensing, LLC
- Applicants: Microsoft Technology Licensing, LLC
- Inventors: Huichao Xue
- Web page: https://patentgrep.com/patent/US20210064684A1

## CPC

- G06F 16/9538: Presentation of query results
- G06F 16/9035: Filtering based on additional data, e.g. user or group profiles
- G06F 16/9535: Search customisation based on user profiles and personalisation
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Methods, systems, and computer programs are presented for search optimization based on relevant-parameter selection. One method includes an operation for training a machine-learning program with information about users of an online service to generate a machine-learning model that calculates parameter preference scores for a plurality of parameters. Further, the method includes operations for detecting a job search for a user, identifying user parameters associated with the user, and calculating, by the machine-learning model, the parameter preference scores for the user parameters. Further, search parameters are determined by selecting a predetermined number of user parameters base on the parameter preference scores. A search of a job-postings database is performed with the search parameters, and the results are presented on a display.

## Claims

### Claim 1

A method comprising:
training, by one or a ore processors, a machine-learning program with information about users of an online service to generate a machine-learning model that calculates parameter preference scores for a plurality of parameters;
detecting, by the one or more processors, a job search for a user;
identifying, by the one or more processors, user parameters associated with the user;
calculating, by the machine-learning model, the parameter preference scores for the user parameters;
determining, by the one or more processors, search parameters by selecting a predetermined number of user parameters base on the parameter preference scores;
performing, by the one or more processors, a search of a job-postings database with the search parameters; and
causing, by the one or more processors, presentation on a display of job postings based on the performed search of the job-postings database.

### Claim 2 (depends on 1)

The method as recited in claim 1, wherein the parameter preference score is a probability of the parameter being a preferred parameter for the user.

### Claim 3 (depends on 1)

The method as recited in claim 1, wherein determining the search parameters further comprises:
determining, for each user parameter, a number of job postings that match the user parameter;
calculating, for each user parameter, an average parameter relevance by dividing the parameter preference score by the number of job postings matching the user parameter; and
selecting the search parameters based on the average parameter relevance.

### Claim 4 (depends on 1)

The method as recited in claim 1, wherein the machine-learning program is trained with information about users of the online service, the information comprising job applications submitted by the users of the online service, and user profile information.

### Claim 5 (depends on 1)

The method as recited in claim 1, further comprising:
before presenting the job postings, filtering results of the search of the job-postings database.

### Claim 6 (depends on 5)

The method as recited in claim 5, wherein the filtering is performed by another machine-learning model based on the user information and information of the job postings.

### Claim 7 (depends on 1)

The method as recited in claim 1, wherein the search of the job-postings database comprises a database query that includes a logical AND of matching the selected parameters.

### Claim 8 (depends on 1)

The method as recited in claim 1, wherein identifying the user parameters comprises:
determining which from the plurality of parameters matches info, enation about the user.

### Claim 9 (depends on 1)

The method as recited in claim 1, wherein the plurality of parameters includes skill parameters, title parameters, and functional area parameters.

### Claim 10 (depends on 1)

The method as recited in claim 1, detecting the job search for the user comprises:
creating the job search, without user explicit request, to determine possible job postings that match profile information of the user.

### Claim 11

A system comprising:
a memory comprising instructions; and
one or more computer processors, wherein the instructions, when executed by the one or more computer processors, cause the system to perform operations comprising:
training, by one or more processors; a machine-learning program with information about users of an online service to generate a machine-learning model that calculates parameter preference scores for a plurality of parameters;
detecting; by the one or more processors, a job search for a user;
identifying, by the one or more processors, user parameters associated with the user;
calculating, by the machine-learning model, the parameter preference scores for the user parameters;
determining, by the one or more processors, search parameters by selecting a predetermined number of user parameters base on the parameter preference scores;
performing, by the one or more processors, a search of a job-postings database with the search parameters; and
causing, by the one or more processors, presentation on a display of job postings based on the performed search of the job-postings database.

### Claim 12 (depends on 11)

The system as recited in claim 11, wherein the parameter preference score is a probability of the parameter being a preferred parameter for the user.

### Claim 13 (depends on 11)

The system as recited in claim 11, wherein determining the search parameters further comprises:
determining, for each user parameter, a number of job postings that match the user parameter;
calculating, for each user parameter, an average parameter relevance by dividing the parameter preference score by the number of job postings matching the user parameter; and
selecting the search parameters based on the average parameter relevance.

### Claim 14 (depends on 11)

The system as recited in claim 11, wherein the machine-learning program is trained with information about users of the online service, the information comprising job applications submitted by the users of the online service, and user profile information.

### Claim 15 (depends on 11)

The system as recited in claim 11, wherein the instructions further cause the one or more computer processors to perform operations comprising:
before presenting the job postings, filtering results of the search of the job-postings database, wherein the filtering is performed by another machine-learning model based on the user information and information of the job postings.

### Claim 16

A non-transitory machine-readable storage medium including instructions that, when executed by a machine, cause the machine to perform operations comprising:
training, by one or more processors, a machine-learning program with information about users of an online service to generate a machine-learning model that calculates parameter preference scores for a plurality of parameters;
detecting, by the one or more processors, a job search for a user;
identifying, by the one or more processors, user parameters associated with the user;
calculating, by the machine-learning model, the parameter preference scores for the user parameters;
determining, by the one or more processors, search parameters by selecting a predetermined number of user parameters base on the parameter preference scores;
performing, by the one or more processors, a search of a job-postings database with the search parameters; and
causing, by the one or more processors, presentation on a display of job postings based on the performed search of the job-postings database.

### Claim 17 (depends on 16)

The non-transitory machine-readable storage medium as recited in claim 16, wherein the parameter preference score is a probability of the parameter being a preferred parameter for the user.

### Claim 18 (depends on 16)

The non-transitory machine-readable storage medium as recited in claim 16, wherein determining the search parameters further comprises:
determining, for each user parameter, a number of job postings that match the user parameter;
calculating, for each user parameter, an average parameter relevance by dividing the parameter preference score by the number of job postings matching the user parameter; and
selecting the search parameters based on the average parameter relevance.

### Claim 19 (depends on 16)

The non-transitory machine-readable storage medium as recited in claim 16, wherein the machine-learning program is trained with information about users of the online service, the information comprising job applications submitted by the users of the online service; and user profile information.

### Claim 20 (depends on 16)

The non-transitory machine-readable storage medium as recited in claim 16, wherein the machine further performs operations comprising:
before presenting the job postings, filtering results of the search of the job-postings database, wherein the filtering is performed by another machine-learning model based on the user information and information of the job postings.
