# US 12,002,569 B2 · Generating multi-dimensional recommendation data objects based on decentralized crowd sourcing

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

- Number: US 12,002,569 B2
- Document ID: US12002569B2
- Status: Granted
- Kind: B2
- Application number: 17/143,004
- Filed: 2021-01-06
- Published: 2022-07-07
- Granted: 2024-06-04
- Earliest claimed: 2021-01-06
- Assignees: Optum Technology, Inc.
- Applicants: Optum Technology, Inc.
- Inventors: Amit Krishna; Rahul Dutta; Shubhendu Shekhar; Atul Kumar; Love Hasija
- Web page: https://patentgrep.com/patent/US12002569B2

## CPC

- G16H 20/70: relating to mental therapies, e.g. psychological therapy or autogenous training
- G16H 50/30: for calculating health indices; for individual health risk assessment
- G06N 20/00: Machine learning
- G16H 10/20: for electronic clinical trials or questionnaires
- G16H 10/60: for patient-specific data, e.g. for electronic patient records
- G16H 40/63: for local operation
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- H04L 9/0643: {Hash functions, e.g. MD5, SHA, HMAC or f9 MAC}
- H04L 9/3231: {Biological data, e.g. fingerprint, voice or retina (network architectures or network communication protocols for supporting authentication of entities using biometrical features in a packet data network H04L63/0861)}
- H04L 9/50: {using hash chains, e.g. blockchains or hash trees}
- H04L 2209/88: Medical equipments
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Various embodiments provide for decentralized crowd sourced generation of recommendation data objects. An example apparatus receives, originating from an external computing device, a recommendation data object request, the recommendation data object request comprising a user identifier and one or more user attributes. The example apparatus may further retrieve, based on a predictive recommendation model, one or more therapy identifiers associated with a therapy efficacy score exceeding a therapy efficacy score threshold for attributes of a first attributes set associated with a first cluster identifier, the first attributes set comprising one or more of the one or more user attributes. The predictive recommendation model is trained based at least in part on trusted efficacy blocks of a distributed ledger, where the trusted efficacy blocks are added to the distributed ledger when a number of therapy efficacy transaction blocks associated with a given therapy identifier and having a therapy efficacy score exceeding a therapy efficacy score threshold meets or exceeds a therapy efficacy transaction block threshold. The example apparatus may further transmit, to the external computing device, a recommendation data object configured for rendering for display via a display device of the external computing device, where the recommendation data object comprises one or more of the one or more therapy identifiers.

## Claims

### Claim 1

A computing system comprising one or more processors and memory storing instructions that, with the one or more processors, configure the computing system to:
train, using a supervisory training function, a machine learning collaborative filtering model, using (i) a plurality of feature vectors corresponding to a plurality of user identifiers and (ii) known therapy efficacy data, to generate a trained machine learning collaborative filtering model;
receive, originating from an external computing device, a recommendation data object request, the recommendation data object request comprising a user identifier and one or more user attributes;
input the one or more user attributes into the trained machine learning collaborative filtering model;
generate, using the trained machine learning collaborative filtering model, one or more therapy identifiers associated with one or more therapy efficacy scores exceeding a therapy efficacy score threshold for one or more of the one or more user attributes;
provide, to the external computing device and in response to the recommendation data object request, a recommendation data object comprising one or more of the one or more therapy identifiers, causing the external computing device to provide, based at least in part on the recommendation data object, a therapy feedback object for generating a therapy efficacy transaction block for a distributed ledger, wherein the therapy efficacy transaction block comprises a therapy efficacy score for the user identifier and a therapy identifier; and
retrain, using a reinforcement learning function, the trained machine learning collaborative filtering model using (i) the plurality of feature vectors and (ii) a plurality of trusted efficacy blocks of the distributed ledger, wherein the plurality of trusted efficacy blocks is generated by:
generating, based at least in part on a therapy efficacy model, a plurality of therapy efficacy transaction blocks based at least in part on (i) a respective therapy efficacy score and (ii) one or more respective efficacy indication objects associated with each of a plurality of therapy identifiers, and
generating, based at least in part on an efficacy consensus protocol, a trusted efficacy block from the distributed ledger in response to a number of the plurality of therapy efficacy transaction blocks satisfying a therapy efficacy transaction block threshold.

### Claim 2 (depends on 1)

The computing system of claim 1, wherein the distributed ledger comprises a block chain comprising a plurality of transaction blocks.

### Claim 3 (depends on 1)

The computing system of claim 1, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
generate, based at least in part on the user identifier and the one or more user attributes, a first feature vector comprising a plurality of data records, wherein each of the one or more user attributes corresponds to a unique data record of the plurality of data records; and
assign the first feature vector to a first cluster of a plurality of clusters according to a clustering model, the first cluster associated with the one or more of the one or more user attributes.

### Claim 4 (depends on 3)

The computing system of claim 3, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
generate one or more of a biometrics transaction block or a patient profile transaction block based at least in part on the first feature vector; and
add the one or more of the biometrics transaction block or the patient profile transaction block to the distributed ledger.

### Claim 5 (depends on 4)

The computing system of claim 4, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
add the one or more of the biometrics transaction block or the patient profile transaction block to the distributed ledger based at least in part on validating a hash of all blocks in the distributed ledger.

### Claim 6 (depends on 1)

The computing system of claim 1, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
receive one or more therapy feedback objects comprising a plurality of therapy feedback data records and at least one therapy identifier of the one or more therapy identifiers;
generate, based in part on the therapy efficacy model and the plurality of therapy feedback data records, the therapy efficacy score for the at least one therapy identifier;
generate a therapy efficacy transaction block based at least in part on the therapy efficacy score, the at least one therapy identifier, and the one or more therapy feedback objects; and
cause storage of the therapy efficacy transaction block on the distributed ledger.

### Claim 7 (depends on 1)

The computing system of claim 1, wherein the therapy efficacy model is trained based at least in part on historical patient profile data, historical patient biometric data, and known therapy efficacy data.

### Claim 8 (depends on 1)

The computing system of claim 1, wherein the trained machine learning collaborative filtering model is periodically re-trained based at least in part on new trusted efficacy blocks added to the distributed ledger.

### Claim 9 (depends on 1)

The computing system of claim 1, wherein the recommendation data object is associated with a recommendation data object identifier.

### Claim 10 (depends on 1)

The computing system of claim 1, wherein the recommendation data object is associated with a recommendation category.

### Claim 11 (depends on 10)

The computing system of claim 10, wherein the recommendation category is one or more of lifestyle recommendation, mindfulness activity, sleep pattern, exercise regime, workout strategy, behavioral guideline, daily routine adjustment.

### Claim 12 (depends on 1)

The computing system of claim 1, wherein at least one therapy identifier of the one or more therapy identifiers is associated with a recommender category.

### Claim 13 (depends on 12)

The computing system of claim 12, wherein the recommender category is one or more of psychiatrist, behavioral therapist, researcher, educational institute, wellness coach, lifestyle guru, life coach.

### Claim 14 (depends on 3)

The computing system of claim 3, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
generate, based at least in part on the first feature vector, a plurality of other feature vectors associated with a plurality of other user identifiers retrieved from a data repository; and
generate, via the clustering model, a plurality of clusters, wherein each cluster of the plurality of clusters comprises one or more other feature vectors of the plurality of other feature vectors.

### Claim 15 (depends on 14)

The computing system of claim 14, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
generate, based at least in part on similarity scores associated with each cluster of the plurality of clusters, a plurality of sub-chains, wherein each sub-chain of the plurality of sub-chains comprises one or more clusters having similarity scores meeting or exceeding a similarity score threshold.

### Claim 16 (depends on 1)

The computing system of claim 1, wherein the memory stores instructions that, with the one or more processors, further configure the computing system to:
receive, originating from one or more electronic sensors associated with the user identifier, biometric data associated with the user identifier;
generate, based at least in part on the user identifier and the biometric data, a biometric data transaction block; and
add the biometric data transaction block to the distributed ledger.

### Claim 17 (depends on 16)

The computing system of claim 16, wherein the biometric data comprises one or more of physiological or psychological data.

### Claim 18 (depends on 16)

The computing system of claim 16, wherein the one or more electronic sensors comprise one or more of wearable sensors, smart watches, fitness devices, audio recording devices, video recording devices, biometric sensors, throat patches, EEG sensors, galvanic skin reaction sensors, sweat sensors, or blood pressure sensors.

### Claim 19 (depends on 1)

The computing system of claim 1, wherein the therapy efficacy model is trained based at least in part on feature vectors associated with a plurality of control group objects.

### Claim 20 (depends on 19)

The computing system of claim 19, wherein the plurality of control group objects is identified based at least in part on common attributes associated with clusters.

### Claim 21

A method for decentralized crowd sourced generation of recommendation data objects, the method comprising:
training, by one or more processors and using a supervisory training function, a machine learning collaborative filtering model, using (i) a plurality of feature vectors corresponding to a plurality of user identifiers and (ii) known therapy efficacy data, to generate a trained machine learning collaborative filtering model;
receiving, by the one or more processors, a recommendation data object request originating from an external computing device, the recommendation data object request comprising a user identifier and one or more user attributes;
inputting, by the one or more processors, the one or more user attributes into the trained machine learning collaborative filtering model;
generating, by the one or more processors and using the trained machine learning collaborative filtering model, one or more therapy identifiers associated with one or more therapy efficacy scores exceeding a therapy efficacy score threshold for one or more of the one or more user attributes;
providing, by the one or more processors and to the external computing device in response to the recommendation data object request, a recommendation data object comprising one or more of the one or more therapy identifiers, causing the external computing device to provide, based at least in part on the recommendation data object, a therapy feedback object for generating a therapy efficacy transaction block for a distributed ledger, wherein the therapy efficacy transaction block comprises a therapy efficacy score for the user identifier and a therapy identifier; and
retraining, by the one or more processors and using a reinforcement learning function, the trained machine learning collaborative filtering model using (i) the plurality of feature vectors and (ii) a plurality of trusted efficacy blocks of the distributed ledger, wherein the plurality of trusted efficacy blocks is generated by:
generating, based at least in part on a therapy efficacy model, a plurality of therapy efficacy transaction blocks based at least in part on (i) a respective therapy efficacy score and (ii) one or more respective efficacy indication objects associated with each of a plurality of therapy identifiers, and
generating, based at least in part on an efficacy consensus protocol, a trusted efficacy block from the distributed ledger in response to a number of the plurality of therapy efficacy transaction blocks satisfying a therapy efficacy transaction block threshold.

### Claim 22

A computer-readable medium comprising at least one non-transitory storage medium storing instructions that, when executed by one or more processors, cause an apparatus to:
train, using a supervisory training function, a machine learning collaborative filtering model, using (i) a plurality of feature vectors corresponding to a plurality of user identifiers and (ii) known therapy efficacy data, to generate a trained machine learning collaborative filtering model;
receive, originating from an external computing device, a recommendation data object request, the recommendation data object request comprising a user identifier and one or more user attributes;
input the one or more user attributes into the trained machine learning collaborative filtering model;
generate, using the trained machine learning collaborative filtering model, one or more therapy identifiers associated with one or more therapy efficacy scores exceeding a therapy efficacy score threshold for one or more of the one or more user attributes;
provide, to the external computing device and in response to the recommendation data object request, a recommendation data object comprising one or more of the one or more therapy identifiers, causing the external computing device to provide, based at least in part on the recommendation data object, a therapy feedback object for generating a therapy efficacy transaction block for a distributed ledger, wherein the therapy efficacy transaction block comprises a therapy efficacy score for the user identifier and a therapy identifier; and
retrain, using a reinforcement learning function, the trained machine learning collaborative filtering model using (i) the plurality of feature vectors and (ii) a plurality of trusted efficacy blocks of the distributed ledger, wherein the plurality of trusted efficacy blocks is generated by:
generating, based at least in part on a therapy efficacy model, a plurality of therapy efficacy transaction blocks based at least in part on (i) a respective therapy efficacy score and (ii) one or more respective efficacy indication objects associated with each of a plurality of therapy identifiers, and
generating, based at least in part on an efficacy consensus protocol, a trusted efficacy block from the distributed ledger in response to a number of the plurality of therapy efficacy transaction blocks satisfying a therapy efficacy transaction block threshold.