# US 2023/0054624 A1 · Dynamic detection of cross-document associations

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

- Number: US 2023/0054624 A1
- Document ID: US20230054624A1
- Status: Published application
- Kind: A1
- Application number: 18/048,074
- Filed: 2022-10-20
- Published: 2023-02-23
- Earliest claimed: 2020-09-09
- Assignees: Optum Technology, Inc.
- Applicants: Optum Technology, Inc.
- Inventors: Swapna Sourav Rout; Sharlene L. Tan; Vamsi Bhandaru; Vineet Shukla; Akshaya D. N; Kartik Chaudhary
- Web page: https://patentgrep.com/patent/US20230054624A1

## CPC

- G06F 40/295: Named entity recognition
- G06V 30/40: Document-oriented image-based pattern recognition
- G06F 16/316: {Indexing structures}
- G06F 16/338: Presentation of query results
- G06F 16/382: {using citations (hypermedia G06F16/94)}
- G06F 16/383: using metadata automatically derived from the content
- G06F 17/15: Correlation function computation {including computation of convolution operations (arithmetic circuits for sum of products per se, e.g. multiply-accumulators G06F7/5443; digital filters, e.g. FIR, IIR, adaptive filters H03H17/00)}
- G06F 40/216: using statistical methods
- G06F 40/30: Semantic analysis
- G06N 20/00: Machine learning
- G06Q 10/04: Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202)
- G06Q 40/08: Insurance
- G06V 20/20: in augmented reality scenes
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- 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

## Abstract

Systems, methods, and computer program products may be configured to generate a set of related document objects for a predictive entity and/or to generate an optimal document sequence for a set of related document objects. In one embodiment, for example, a set of related document objects for a predictive entity is generated by processing entity metadata features associated with the predictive entity using an entity-document correlation machine learning model, and an optimal document sequence is generated for the set of related document objects by processing the set of related document objects using a document sequence optimization machine learning model.

## Claims

### Claim 1

A computer-implemented method for generating a set of related document objects for a predictive entity, the computer-implemented method comprising:
identifying, by one or more processors, a set of entity metadata features associated with the predictive entity;
generating, by the one or more processors and using an entity-document correlation machine learning model and based at least in part on the set of entity metadata features, the set of related document objects for the predictive entity, wherein (i) the entity-document correlation machine learning model is generated at least in part using a set of inferred annotative relationships between a set of input index documents and a set of inferred reference documents, (ii) each inferred annotative relationship of the set of inferred annotative relationships is associated with a corresponding input index document of the set of input index documents, and (iii) each inferred annotative relationship of the set of inferred annotative relationships describes a related reference document subset of the set of inferred reference documents for the corresponding input index document that is associated with the inferred annotative relationship; and
initiating, by the one or more processors, the performance of one or more prediction-based actions based at least in part on the set of related document objects.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein generating the entity-document correlation machine learning model comprises:
identifying a set of prior annotative relationships between a set of prior index documents and a set of prior reference documents;
generating a cross-document annotation machine learning model based at least in part on the set of prior annotative relationships, wherein the cross-document annotation machine learning model is configured to process each corresponding input index document in the set of input index documents to generate the related reference document subset for the corresponding input index document;
for each corresponding input index document of the set of inferred reference documents:
processing the corresponding input index document using the cross-document annotation machine learning model to generate the related reference document subset for the corresponding input index document, and
determining a set of input index document entity metadata features for the corresponding input index document; and
generating the entity-document correlation machine learning model based at least in part on each related reference document subset for the set of input index documents and each set of input index document entity metadata features for the set of input index documents.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, wherein:
the entity-document correlation machine learning model is further configured to generate a related document sequence for the set of related document objects, and
the related document sequence defines a related document ordering value for each related document objects in the set of related document objects.

### Claim 4 (depends on 3)

The computer-implemented method of claim 3, wherein:
each inferred annotative relationship of the set of inferred annotative relationships defines a related reference document sequence for the inferred reference document subset that is associated with the inferred annotative relationship;
the related reference document sequence for an inferred reference document subset defines a reference document ordering value for each inferred reference document in the inferred reference document subset; and
the entity-document correlation machine learning model is trained to generate the related document sequence based at least in part on each related reference document sequence defined by the set of inferred annotative relationships.

### Claim 5 (depends on 4)

The computer-implemented method of claim 4, wherein:
each inferred annotative relationship of the set of inferred annotative relationships is associated with one or more cross-document annotations,
each inferred annotative relationship of the set of inferred annotative relationships is associated with a corresponding text segment of one or more text segments of the corresponding input index document associated with the inferred annotative relationship,
for each corresponding input text document of the set of input text documents, the one or more text segments for the corresponding input index document are associated with a text segment order, and
each related reference document sequence defined by the set of inferred annotative relationships is determined based at least in part on the text segment order of the corresponding input index document that is associated with the inferred annotative relationship.

### Claim 6 (depends on 1)

The computer-implemented method of claim 1, wherein the set of entity metadata features associated with the predictive entity comprise one or more entity demographic features associated with the predictive entity and one or more diagnostic features associated with the predictive entity.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6, wherein the set of entity metadata features associated with the predictive entity comprise one or more policy-related features associated with the predictive entity.

### Claim 8

An apparatus for generating a subset of related document objects for a predictive entity, the apparatus comprising at least one processor and at least one non-transitory memory including program code, the at least one non-transitory memory and the program code configured to, with the at least one processor, cause the apparatus to at least:
identify a set of entity metadata features associated with the predictive entity;
generate, using an entity-document correlation machine learning model and based at least in part on the set of entity metadata features, the set of related document objects for the predictive entity, wherein (i) the entity-document correlation machine learning model is generated at least in part using a set of inferred annotative relationships between a set of input index documents and a set of inferred reference documents, (ii) each inferred annotative relationship of the set of inferred annotative relationships is associated with a corresponding input index document of the set of input index documents, and (iii) each inferred annotative relationship of the set of inferred annotative relationships describes a related reference document subset of the set of inferred reference documents for the corresponding input index document that is associated with the inferred annotative relationship; and
initiate the performance of one or more prediction-based actions based at least in part on the set of related document objects.

### Claim 9 (depends on 8)

The apparatus of claim 8, wherein generating the entity-document correlation machine learning model comprises:
identifying a set of prior annotative relationships between a set of prior index documents and a set of prior reference documents;
generating a cross-document annotation machine learning model based at least in part on the set of prior annotative relationships, wherein the cross-document annotation machine learning model is configured to process each corresponding input index document in the set of input index documents to generate the related reference document subset for the corresponding input index document;
for each corresponding input index document of the set of inferred reference documents:
processing the corresponding input index document using the cross-document annotation machine learning model to generate the related reference document subset for the corresponding input index document, and
determining a set of input index document entity metadata features for the corresponding input index document; and
generating the entity-document correlation machine learning model based at least in part on each related reference document subset for the set of input index documents and each set of input index document entity metadata features for the set of input index documents.

### Claim 10 (depends on 8)

The apparatus of claim 8, wherein:
the entity-document correlation machine learning model is further configured to generate a related document sequence for the set of related document objects, and
the related document sequence defines a related document ordering value for each related document objects in the set of related document objects.

### Claim 11 (depends on 10)

The apparatus of claim 10, wherein:
each inferred annotative relationship of the set of inferred annotative relationships defines a related reference document sequence for the inferred reference document subset that is associated with the inferred annotative relationship;
the related reference document sequence for an inferred reference document subset defines a reference document ordering value for each inferred reference document in the inferred reference document subset; and
the entity-document correlation machine learning model is trained to generate the related document sequence based at least in part on each related reference document sequence defined by the set of inferred annotative relationships.

### Claim 12 (depends on 11)

The apparatus of claim 11, wherein:
each inferred annotative relationship of the set of inferred annotative relationships describes one or more cross-document annotations,
each inferred annotative relationship of the set of inferred annotative relationships is associated with a corresponding text segment of one or more text segments of the corresponding input index document associated with the inferred annotative relationship,
for each corresponding input text document of the set of input text documents, the one or more text segments for the corresponding input index document are associated with a text segment order, and
each related reference document sequence defined by the set of inferred annotative relationships is determined based at least in part on the text segment order of the corresponding input index document that is associated with the inferred annotative relationship.

### Claim 13 (depends on 8)

The apparatus of claim 8, wherein the set of entity metadata features associated with the predictive entity comprise one or more entity demographic features associated with the predictive entity and one or more diagnostic features associated with the predictive entity.

### Claim 14 (depends on 13)

The apparatus of claim 13, wherein the set of entity metadata features associated with the predictive entity comprise one or more policy-related features associated with the predictive entity.

### Claim 15

A non-transitory computer storage medium comprising instructions for generating a subset of related document objects for a predictive entity, the instructions being configured to cause one or more processors to at least perform operations configured to:
identify a set of entity metadata features associated with the predictive entity;
generate, using an entity-document correlation machine learning model and based at least in part on the set of entity metadata features, the set of related document objects for the predictive entity, wherein (i) the entity-document correlation machine learning model is generated at least in part using a set of inferred annotative relationships between a set of input index documents and a set of inferred reference documents, (ii) each inferred annotative relationship of the set of inferred annotative relationships is associated with a corresponding input index document of the set of input index documents, and (iii) each inferred annotative relationship of the set of inferred annotative relationships describes a related reference document subset of the set of inferred reference documents for the corresponding input index document that is associated with the inferred annotative relationship; and
initiate the performance of one or more prediction-based actions based at least in part on the set of related document objects.

### Claim 16 (depends on 15)

The non-transitory computer storage medium of claim 15, wherein generating the entity-document correlation machine learning model comprises:
identifying a set of prior annotative relationships between a set of prior index documents and a set of prior reference documents;
generating a cross-document annotation machine learning model based at least in part on the set of prior annotative relationships, wherein the cross-document annotation machine learning model is configured to process each corresponding input index document in the set of input index documents to generate the related reference document subset for the corresponding input index document;
for each corresponding input index document of the set of inferred reference documents:
processing the corresponding input index document using the cross-document annotation machine learning model to generate the related reference document subset for the corresponding input index document, and
determining a set of input index document entity metadata features for the corresponding input index document; and
generating the entity-document correlation machine learning model based at least in part on each related reference document subset for the set of input index documents and each set of input index document entity metadata features for the set of input index documents.

### Claim 17 (depends on 15)

The non-transitory computer storage medium of claim 15, wherein:
the entity-document correlation machine learning model is further configured to generate a related document sequence for the set of related document objects, and
the related document sequence defines a related document ordering value for each related document objects in the set of related document objects.

### Claim 18 (depends on 17)

The non-transitory computer storage medium of claim 17, wherein:
each inferred annotative relationship of the set of inferred annotative relationships defines a related reference document sequence for the inferred reference document subset that is associated with the inferred annotative relationship;
the related reference document sequence for an inferred reference document subset defines a reference document ordering value for each inferred reference document in the inferred reference document subset; and
the entity-document correlation machine learning model is trained to generate the related document sequence based at least in part on each related reference document sequence defined by the set of inferred annotative relationships.

### Claim 19 (depends on 18)

The non-transitory computer storage medium of claim 18, wherein:
each inferred annotative relationship of the set of inferred annotative relationships describes one or more cross-document annotations,
each inferred annotative relationship of the set of inferred annotative relationships describes is associated with a corresponding text segment of one or more text segments of the corresponding input index document associated with the inferred annotative relationship,
for each corresponding input text document of the set of input text documents, the one or more text segments for the corresponding input index document are associated with a text segment order, and
each related reference document sequence defined by the set of inferred annotative relationships is determined based at least in part on the text segment order of the corresponding input index document that is associated with the inferred annotative relationship.

### Claim 20 (depends on 15)

The non-transitory computer storage medium of claim 15, wherein the set of entity metadata features associated with the predictive entity comprise one or more entity demographic features associated with the predictive entity and one or more diagnostic features associated with the predictive entity.
