# US 2023/0054924 A1 · Methods and systems for determining provenance and identity of digital advertising requests solicited by publishers and intermediaries representing publishers

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

- Number: US 2023/0054924 A1
- Document ID: US20230054924A1
- Status: Published application
- Kind: A1
- Application number: 17/980,891
- Filed: 2022-11-04
- Published: 2023-02-23
- Earliest claimed: 2020-08-05
- Assignees: MadHive, Inc.
- Applicants: Madhive, Inc.
- Inventors: Aaron Brown; Tom Bollich; Adam Helfgott
- Web page: https://patentgrep.com/patent/US20230054924A1

## CPC

- G06Q 30/0248: {Avoiding fraud}
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/0475: Generative networks
- G06N 3/0495: Quantised networks; Sparse networks; Compressed networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 3/091: Active learning
- G06Q 30/0264: {based upon schedule}
- 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
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Described is a computer implemented method for determining provenance and identity of a digital advertising request for an advertisement solicited by at least one of a publisher and an intermediary representing the publisher. The method includes receiving, with a transceiver of a first computing device, a first message having first message metrics associated with a candidate entity for identification. The first message is processed, with a processor of the first computing device, to identify a first portion of a candidate entity identity based on the first message metrics. An output related to confidence of the candidate entity identity is produced, with the processor, based on the first portion of the candidate entity identity.

## Claims

### Claim 1

A computer implemented method for determining provenance and using machine learning to generate a digital identity of a first entity purporting to be a second entity defined by one of a publisher and an intermediary representing the publisher, the computer implemented method comprising:
identifying a plurality of metrics of a digital advertising request that is received, using a transceiver of a computing device, from a first entity computing device of the first entity, wherein the plurality of metrics comprises at least one of (i) a serving constraint having a serving constraint value and (ii) a protocol having a protocol value;
converting the plurality of metrics to a plurality of embeddings;
generating, with a processor of the computing device using machine learning, a first portion of the digital identity of the first entity based on the plurality of embeddings;
comparing, with the processor using machine learning, the first portion of the digital identity to a second digital identity associated with the second entity that the first entity purports to be;
determining at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request; and
generating a response to the digital advertising request comprising a solicitation for a reply message, from the first entity, comprising at least one second metric related to at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request.

### Claim 2 (depends on 1)

The computer implemented method of claim 1 further comprising one of (i) sending the response with the transceiver to the first entity computing device and (ii) sending the response with the transceiver to the first entity computing device after a delay.

### Claim 3 (depends on 1)

The computer implemented method of claim 1 further comprising:
after determining the at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request, then processing, with the processor, the first portion of the digital identity using machine learning and at least one clustering algorithm to predict one of (i) a predicted protocol value and (ii) a predicted serving constraint value for the at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request.

### Claim 4 (depends on 1)

The computer implemented method of claim 1 further comprising:
after determining the at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request, then calculating, with the processor using machine learning and at least one clustering algorithm, an instability exhibited by a predicted value relative to the first portion of the digital identity and the second digital identity; and
wherein the predicted value comprises one of a predicted serving constraint value and a predicted protocol value for the at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request.

### Claim 5 (depends on 1)

The computer implemented method of claim 1 further comprising:
generating, with the processor using machine learning and at least one clustering algorithm, a level of confidence that the first entity is the second entity that it purports to be.

### Claim 6 (depends on 1)

The computer implemented method of claim 1 further comprising:
after generating the response to the digital advertising request, then receiving, with the transceiver of the computing device, from the first entity computing device, the reply message comprising the at least one second metric.

### Claim 7 (depends on 6)

The computer implemented method of claim 6, wherein after receiving the reply message, the computer implemented method further comprises:
processing the at least one second metric using machine learning and at least one clustering algorithm to generate a second portion of the digital identity.

### Claim 8 (depends on 7)

The computer implemented method of claim 7, comprising
generating the digital identity by combining, with the processor using machine learning, the first portion of the digital identity and the second portion of the digital identity.

### Claim 9 (depends on 8)

The computer implemented method of claim 8, comprising determining, with the processor using machine learning and the at least one clustering algorithm, a level of confidence that the first entity is the second entity that it purports to be by comparing the digital identity to the second digital identity.

### Claim 10

A computer implemented method for determining provenance and using machine learning to generate a digital identity of a first entity purporting to be a second entity defined by one of a of a publisher and an intermediary representing the publisher, the computer implemented method comprising:
identifying a plurality of metrics of a digital advertising request that is received, using a transceiver of a computing device, from a first entity computing device of the first entity, wherein the plurality of metrics comprises at least one of (i) a serving constraint having a serving constraint value and (ii) a protocol having a protocol value;
converting the plurality of metrics to a plurality of embeddings;
generating, with a processor using machine learning, a first portion of the digital identity associated with the first entity based on the plurality of embeddings;
comparing, with the processor using machine learning, the first portion of the digital identity to a second digital identity associated with the second entity; and
generating, with the processor using machine learning, a level of confidence that the first entity is the second entity that it purports to be.

### Claim 11 (depends on 10)

The computer implemented method of claim 10 further comprising:
processing, with the processor, the first portion of the digital identity using machine learning and at least one clustering algorithm to predict one of (i) a predicted protocol value and (ii) a predicted serving constraint value for at least one second digital identity metric of the second digital identity, wherein the at least one second digital identity metric is not identified in the digital advertising request.

### Claim 12 (depends on 10)

The computer implemented method of claim 10 further comprising:
wherein generating the level of confidence comprises calculating, with the processor using machine learning and at least one clustering algorithm, an instability exhibited by a predicted value relative to the first portion of the digital identity and the second digital identity; and
wherein the predicted value comprises one of a predicted serving constraint value and a predicted protocol value for at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request.

### Claim 13 (depends on 10)

The computer implemented method of claim 10 further comprising determining at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request.

### Claim 14 (depends on 10)

The computer implemented method of claim 10 further comprising generating a response to the digital advertising request comprising a solicitation for a reply message, from the first entity, comprising at least one second metric related to at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request.

### Claim 15 (depends on 10)

The computer implemented method of claim 10 further comprising
generating a response to the digital advertising request comprising a solicitation for a reply message, from the first entity, comprising at least one second metric related to at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request; and
one of (i) sending the response with the transceiver to the first entity computing device and (ii) sending the response with the transceiver to the first entity computing device after a delay.

### Claim 16 (depends on 15)

The computer implemented method of claim 15 further comprising receiving, with the transceiver, from the first entity computing device, the reply message comprising the at least one second metric.

### Claim 17

A computer implemented method for determining provenance and using machine learning to generate a digital identity of a first entity purporting to be a second entity defined by one of a publisher and an intermediary representing the publisher, the computer implemented method comprising:
identifying a plurality of metrics of a digital advertising request that is received, using a transceiver of a computing device, from a first entity computing device of the first entity, wherein the plurality of metrics comprises at least one of (i) a serving constraint having a serving constraint value and (ii) a protocol having a protocol value;
converting the plurality of metrics to a plurality of embeddings;
generating, with a processor using machine learning, a first portion of the digital identity associated with the first entity based on the plurality of embeddings;
comparing, with the processor using machine learning, the first portion of the digital identity to a second digital identity associated with the second entity that the first entity purports to be;
generating a response to the digital advertising request comprising a solicitation for a reply message, from the first entity, comprising at least one second metric related to at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request;
generating, with the processor using machine learning, a second portion of the digital identity of the first entity based on the reply message; and
generating, with the processor using machine learning, the digital identity comprising the first portion and the second portion of the digital identity.

### Claim 18 (depends on 17)

The computer implemented method of claim 17 further comprising, before generating the response, determining the at least one second digital identity metric of the second digital identity that is not identified in the digital advertising request using machine learning and at least one clustering algorithm.

### Claim 19 (depends on 18)

The computer implemented method of claim 18, wherein after generating the response, the computer implemented method further comprises one of (i) sending the response with the transceiver to the first entity computing device and (ii) sending the response with the transceiver to the first entity computing device after a delay.

### Claim 20 (depends on 19)

The computer implemented method of claim 19, wherein generating the second portion of the digital identity comprises:
receiving, with the transceiver, from the first entity computing device, the reply message comprising the at least one second metric; and
processing the at least one second metric using machine learning and the at least one clustering algorithm to generate the second portion of the digital identity.

### Claim 21

A computer implemented method for determining provenance and using machine learning to generate a digital identity of a first entity purporting to be a second entity defined by one of a publisher and an intermediary representing the publisher, the computer implemented method comprising:
identifying a plurality of metrics of a digital advertising request that is received, using a transceiver of a computing device, from a first entity computing device of the first entity;
processing the plurality of metrics using at least one clustering algorithm to identify a second entity metric of a second entity digital identity associated with the second entity that is not identified in the plurality of metrics of the digital advertising request; and
generating, using the at least one clustering algorithm, a level of confidence that the first entity is the second entity.