# US 12,118,040 B1 · Systems and methods for generating synthetic queries

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

- Number: US 12,118,040 B1
- Document ID: US12118040B1
- Status: Granted
- Kind: B1
- Application number: 18/298,988
- Filed: 2023-04-11
- Published: 2024-10-17
- Granted: 2024-10-15
- Earliest claimed: 2023-04-11
- Assignees: Spotify AB
- Applicants: Spotify AB
- Inventors: Gustavo Penha; Enrico Palumbo; Maryam Aziz; Alice Wang; Hugues Bouchard
- Web page: https://patentgrep.com/patent/US12118040B1

## CPC

- G06F 16/90324: {using system suggestions}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

An electronic device provides, to a model trained to generate synthetic queries, information corresponding to a first media content item and a first value for a parameter indicative of query specificity. The device receives, from the model trained to generate synthetic queries, a first query corresponding to the first media content item, wherein the first query is generated in accordance with the first value for the parameter indicative of query specificity. The device provides, to a client device for display at the client device, the first query and receives user selection of the first query. In response to the user selection of the first query, the device provides a first result set of media content items for the first query.

## Claims

### Claim 1

A method, comprising:
providing, to a model trained to generate synthetic queries:
information corresponding to a first media content item; and
a first value for a parameter indicative of query specificity that specifies a broadness and/or narrowness for a first generated synthetic query;
receiving, from the model trained to generate synthetic queries, the first generated synthetic query, wherein the first generated synthetic query corresponds to the first media content item, wherein the first generated synthetic query is generated in accordance with the first value for the parameter indicative of query specificity that specifies a broadness and/or narrowness for the first generated synthetic query;
providing, to a client device for display at the client device, the first generated synthetic query;
receiving user selection of the first generated synthetic query; and
in response to the user selection of the first generated synthetic query, providing a first result set of media content items for the first generated synthetic query.

### Claim 2 (depends on 1)

The method of claim 1, further including, prior to providing the information corresponding to the first media content item to the model trained to generate synthetic queries:
receiving, from the client device, a second query input by a user of the client device; and
determining a second result set of media content items for the second query that includes the first media content item;
wherein the information corresponding to the first media content item is provided to the model trained to generate synthetic queries in accordance with the second result set of media content items for the second query including the first media content item.

### Claim 3 (depends on 1)

The method of claim 1, wherein the information corresponding to the first media content item includes metadata for the first media content item.

### Claim 4 (depends on 1)

The method of claim 1, further comprising, prior to providing the information corresponding to the first media content item to the model trained to generate synthetic queries:
generating, without human labeling, training data for the model; and
training the model to generate synthetic queries using the training data for the model.

### Claim 5 (depends on 4)

The method of claim 4, wherein generating, without human labeling, the training data for the model includes:
obtaining information corresponding to a second media content item;
generating a first training query for the second media content item by applying, to the information corresponding to the second media content item, a first set of rules for generating training queries having a first training value of the parameter indicative of query specificity; and
generating a second training query for the second media content item by applying, to the information corresponding to the second media content item, a second set of rules for generating training queries having a second training value of the parameter indicative of query specificity.

### Claim 6 (depends on 5)

The method of claim 5, wherein training the model to generate synthetic queries using the training data for the model includes training the model using:
the first training query and the first training value; and
the second training query and the second training value.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
providing, to the model trained to generate synthetic queries:
information corresponding to a second media content item; and
a second value for a parameter indicative of query specificity;
receiving, from the model trained to generate synthetic queries, a second generated synthetic query corresponding to the second media content item, wherein the second generated synthetic query is generated in accordance with the second value for the parameter indicative of query specificity; and
training a search retrieval model using the second generated synthetic query and the second media content item as labeled training data.

### Claim 8

An electronic device, comprising:
one or more processors; and
memory storing one or more programs, the one or more programs including instructions for:
providing, to a model trained to generate synthetic queries:
information corresponding to a first media content item; and
a first value for a parameter indicative of query specificity that specifies a broadness and/or narrowness for a first generated synthetic query;
receiving, from the model trained to generate synthetic queries, the first generated synthetic query, wherein the first generated synthetic query corresponds to the first media content item, wherein the first generated synthetic query is generated in accordance with the first value for the parameter indicative of query specificity that specifies a broadness and/or narrowness for the first generated synthetic query;
providing, to a client device for display at the client device, the first generated synthetic query;
receiving user selection of the first generated synthetic query; and
in response to the user selection of the first generated synthetic query, providing a first result set of media content items for the first generated synthetic query.

### Claim 9 (depends on 8)

The electronic device of claim 8, the one or more programs further including instructions for, prior to providing the information corresponding to the first media content item to the model trained to generate synthetic queries:
receiving, from the client device, a second query input by a user of the client device; and
determining a second result set of media content items for the second query that includes the first media content item;
wherein the information corresponding to the first media content item is provided to the model trained to generate synthetic queries in accordance with the second result set of media content items for the second query including the first media content item.

### Claim 10 (depends on 8)

The electronic device of claim 8, wherein the information corresponding to the first media content item includes metadata for the first media content item.

### Claim 11 (depends on 8)

The electronic device of claim 8, the one or more programs further including instructions for, prior to providing the information corresponding to the first media content item to the model trained to generate synthetic queries:
generating, without human labeling, training data for the model; and
training the model to generate synthetic queries using the training data for the model.

### Claim 12 (depends on 11)

The electronic device of claim 11, wherein generating, without human labeling, the training data for the model includes:
obtaining information corresponding to a second media content item;
generating a first training query for the second media content item by applying, to the information corresponding to the second media content item, a first set of rules for generating training queries having a first training value of the parameter indicative of query specificity; and
generating a second training query for the second media content item by applying, to the information corresponding to the second media content item, a second set of rules for generating training queries having a second training value of the parameter indicative of query specificity.

### Claim 13 (depends on 12)

The electronic device of claim 12, wherein training the model to generate synthetic queries using the training data for the model includes training the model using:
the first training query and the first training value; and
the second training query and the second training value.

### Claim 14 (depends on 8)

The electronic device of claim 8, the one or more programs further including instructions for:
providing, to the model trained to generate synthetic queries:
information corresponding to a second media content item; and
a second value for a parameter indicative of query specificity;
receiving, from the model trained to generate synthetic queries, a second generated synthetic query corresponding to the second media content item, wherein the second generated synthetic query is generated in accordance with the second value for the parameter indicative of query specificity; and
training a search retrieval model using the second generated synthetic query and the second media content item as labeled training data.

### Claim 15

A non-transitory computer-readable storage medium storing one or more programs for execution by an electronic device with one or more processors, the one or more programs including instructions for:
providing, to a model trained to generate synthetic queries:
information corresponding to a first media content item; and
a first value for a parameter indicative of query specificity that specifies a broadness and/or narrowness for a first generated synthetic query;
receiving, from the model trained to generate synthetic queries, the first generated synthetic query, wherein the first generated synthetic query corresponds to the first media content item, wherein the first generated synthetic query is generated in accordance with the first value for the parameter indicative of query specificity that specifies a broadness and/or narrowness for the first generated synthetic query;
providing, to a client device for display at the client device, the first generated synthetic query;
receiving user selection of the first generated synthetic query; and
in response to the user selection of the first generated synthetic query, providing a first result set of media content items for the first generated synthetic query.

### Claim 16 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, the one or more programs further including instructions for, prior to providing the information corresponding to the first media content item to the model trained to generate synthetic queries:
receiving, from the client device, a second query input by a user of the client device; and
determining a second result set of media content items for the second query that includes the first media content item;
wherein the information corresponding to the first media content item is provided to the model trained to generate synthetic queries in accordance with the second result set of media content items for the second query including the first media content item.

### Claim 17 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, wherein the information corresponding to the first media content item includes metadata for the first media content item.

### Claim 18 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, the one or more programs further including instructions for, prior to providing the information corresponding to the first media content item to the model trained to generate synthetic queries:
generating, without human labeling, training data for the model; and
training the model to generate synthetic queries using the training data for the model.

### Claim 19 (depends on 18)

The non-transitory computer-readable storage medium of claim 18, wherein generating, without human labeling, the training data for the model includes:
obtaining information corresponding to a second media content item;
generating a first training query for the second media content item by applying, to the information corresponding to the second media content item, a first set of rules for generating training queries having a first training value of the parameter indicative of query specificity; and
generating a second training query for the second media content item by applying, to the information corresponding to the second media content item, a second set of rules for generating training queries having a second training value of the parameter indicative of query specificity.

### Claim 20 (depends on 19)

The non-transitory computer-readable storage medium of claim 19, wherein training the model to generate synthetic queries using the training data for the model includes training the model using:
the first training query and the first training value; and
the second training query and the second training value.