# US 12,602,551 B2 · Generation of synthetic documents for data augmentation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,602,551 B2 - Document ID: US12602551B2 - Status: Granted - Kind: B2 - Application number: 18/401,768 - Filed: 2024-01-02 - Published: 2025-07-03 - Granted: 2026-04-14 - Earliest claimed: 2024-01-02 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Parijat Dube; Vatche Isahagian; Prabhat Maddikunta Reddy - Web page: https://patentgrep.com/patent/US12602551B2 ## CPC - G06F 40/40: Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30) - G06F 40/205: Parsing - G06F 40/295: Named entity recognition - G06F 40/30: Semantic analysis - G06V 30/10: Character recognition - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract According to one embodiment, a method, computer system, and computer program product for generating synthetic business documents for data augmentation is provided. The embodiment may include identifying a first set of key value pairs (KVPs) within a set of business documents spanning multiple domains. The embodiment may include creating domain-agnostic models of spatial distribution and content distribution of KVPs within the set. The embodiment may include grounding the domain-agnostic models of spatial distribution and content distribution using a second set of domain-specific KVPs to derive domain-specific models of spatial distribution and content distribution of KVPs within the second set. The embodiment may include generating a set of synthetic domain-specific business documents using the derived domain-specific models of spatial distribution and content distribution. The embodiment may include augmenting a training data set of a large language model (LLM) with the set of synthetic domain-specific business documents. ## Claims ### Claim 1 A computer-implemented method, the method comprising: identifying a first set of key value pairs (KVPs) within a set of business documents spanning multiple domains; creating a domain-agnostic model of spatial distribution of KVPs within the set; creating a domain-agnostic model of content distribution of KVPs within the set; grounding the domain-agnostic model of spatial distribution and the domain-agnostic model of content distribution using a second set of domain-specific KVPs to derive domain-specific models of spatial distribution and content distribution of KVPs within the second set; generating a set of synthetic domain-specific business documents using the derived domain-specific models of spatial distribution and content distribution; augmenting a training data set of a large language model (LLM) with the set of synthetic domain-specific business documents to create an augmented training data set; training the LLM using the augmented training data set; and performing, via the LLM, domain-specific document understanding tasks. ### Claim 2 (depends on 1) The method of claim 1, wherein identifying the first set of KVPs comprises parsing the set of business documents to extract KVPs via a technique selected from the group consisting of optical character recognition (OCR), intelligent character recognition (ICR), and named entity recognition (NER). ### Claim 3 (depends on 1) The method of claim 1, wherein spatial distributions of the KVPs of the first set and the KVPs of the second set are modeled as multivariate Gaussian distributions. ### Claim 4 (depends on 1) The method of claim 1, wherein modeling content distributions of KVPs comprises identifying structural and semantic patterns for displaying values corresponding to different keys. ### Claim 5 (depends on 1) The method of claim 1, wherein the second set of domain-specific KVPs is a subset of the first set of KVPs. ### Claim 6 (depends on 1) The method of claim 1, wherein a document of the set of synthetic domain-specific business documents is generated by sampling locations of different domain-specific KVPs from the derived domain-specific model of spatial distribution such that there is no overlap in placement of the domain-specific KVPs within the document, and wherein a location of a domain-specific KVP is represented by a vector which specifies positional coordinates of a bounding box which corresponds to the domain-specific KVP. ### Claim 7 A computer system, the computer system comprising: one or more processors, one or more computer-readable memories, one or more computer-readable storage medium, and program instructions stored on at least one of the one or more computer-readable storage medium for execution by at least one of the one or more processors via at least one of the one or more computer-readable memories, wherein the computer system is capable of performing a method comprising: identifying a first set of key value pairs (KVPs) within a set of business documents spanning multiple domains; creating a domain-agnostic model of spatial distribution of KVPs within the set; creating a domain-agnostic model of content distribution of KVPs within the set; grounding the domain-agnostic model of spatial distribution and the domain-agnostic model of content distribution using a second set of domain-specific KVPs to derive domain-specific models of spatial distribution and content distribution of KVPs within the second set; generating a set of synthetic domain-specific business documents using the derived domain-specific models of spatial distribution and content distribution; augmenting a training data set of a large language model (LLM) with the set of synthetic domain-specific business documents to create an augmented training data set; training the LLM using the augmented training data set; and performing, via the LLM, domain-specific document understanding tasks. ### Claim 8 (depends on 7) The computer system of claim 7, wherein identifying the first set of KVPs comprises parsing the set of business documents to extract KVPs via a technique selected from the group consisting of optical character recognition (OCR), intelligent character recognition (ICR), and named entity recognition (NER). ### Claim 9 (depends on 7) The computer system of claim 7, wherein spatial distributions of the KVPs of the first set and the KVPs of the second set are modeled as multivariate Gaussian distributions. ### Claim 10 (depends on 7) The computer system of claim 7, wherein modeling content distributions of KVPs comprises identifying structural and semantic patterns for displaying values corresponding to different keys. ### Claim 11 (depends on 7) The computer system of claim 7, wherein the second set of domain-specific KVPs is a subset of the first set of KVPs. ### Claim 12 (depends on 7) The computer system of claim 7, wherein a document of the set of synthetic domain-specific business documents is generated by sampling locations of different domain-specific KVPs from the derived domain-specific model of spatial distribution such that there is no overlap in placement of the domain-specific KVPs within the document, and wherein a location of a domain-specific KVP is represented by a vector which specifies positional coordinates of a bounding box which corresponds to the domain-specific KVP. ### Claim 13 A computer program product, the computer program product comprising: one or more computer-readable storage medium and program instructions stored on at least one of the one or more computer-readable storage medium, the program instructions executable by a processor capable of performing a method, the method comprising: identifying a first set of key value pairs (KVPs) within a set of business documents spanning multiple domains; creating a domain-agnostic model of spatial distribution of KVPs within the set; creating a domain-agnostic model of content distribution of KVPs within the set; grounding the domain-agnostic model of spatial distribution and the domain-agnostic model of content distribution using a second set of domain-specific KVPs to derive domain-specific models of spatial distribution and content distribution of KVPs within the second set; generating a set of synthetic domain-specific business documents using the derived domain-specific models of spatial distribution and content distribution; augmenting a training data set of a large language model (LLM) with the set of synthetic domain-specific business documents to create an augmented training data set; training the LLM using the augmented training data set; and performing, via the LLM, domain-specific document understanding tasks. ### Claim 14 (depends on 13) The computer program product of claim 13, wherein identifying the first set of KVPs comprises parsing the set of business documents to extract KVPs via a technique selected from the group consisting of optical character recognition (OCR), intelligent character recognition (ICR), and named entity recognition (NER). ### Claim 15 (depends on 13) The computer program product of claim 13, wherein spatial distributions of the KVPs of the first set and the KVPs of the second set are modeled as multivariate Gaussian distributions. ### Claim 16 (depends on 13) The computer program product of claim 13, wherein modeling content distributions of KVPs comprises identifying structural and semantic patterns for displaying values corresponding to different keys. ### Claim 17 (depends on 13) The computer program product of claim 13, wherein the second set of domain-specific KVPs is a subset of the first set of KVPs. ### Claim 18 (depends on 13) The computer program product of claim 13, wherein a document of the set of synthetic domain-specific business documents is generated by sampling locations of different domain-specific KVPs from the derived domain-specific model of spatial distribution such that there is no overlap in placement of the domain-specific KVPs within the document, and wherein a location of a domain-specific KVP is represented by a vector which specifies positional coordinates of a bounding box which corresponds to the domain-specific KVP.