Semantic Annotation Alignment for Multi-Ontology Document Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing natural language processing systems struggle to generate structured semantic annotations from unstructured documents due to the challenge of aligning deciphered key information to a variety of customized ontologies for different use cases, particularly in domains like news and medication.

Innovation Solution

A machine learning-based natural language processing model is trained to extract unstructured data and output structured data using a generic document ontology, followed by aligning these annotations to a target ontology specific to a use case, generating target structured annotations that include entities and relations between them, and integrating them into a knowledge representation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a natural language processing model is trained to extract unstructured data using a generic document ontology, then the model can process multiple domains and use cases, but the extracted structured annotations cannot be directly aligned to specific target ontologies without additional processing

Engineering Contradiction:

Improveability to process multiple domainsVSAvoidalignment process complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces an alignment module as an intermediary component that bridges the generic document ontology and target-specific ontologies. This module automatically aligns generic structured annotations to target ontologies by mapping entities and relations, thereby resolving the contradiction by adding a specialized component that enables multi-domain processing while maintaining manageable complexity through automation.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system employs a universal NLP model trained on a generic document ontology that can process multiple domains and use cases. This universal model serves as a multi-functional foundation, extracting structured annotations that can be adapted to various target ontologies through the alignment process, thus achieving versatility while managing complexity through shared processing capabilities.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Adaptability or versatility

If the system generates generic structured annotations that can be aligned to multiple target ontologies, then the system becomes more versatile, but the process of aligning to specific use cases increases processing time

Engineering Contradiction:

Improvealignment to multiple ontologiesVSAvoidprocessing time for alignment

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs preliminary extraction of structured annotations using a generic document ontology before alignment to target ontologies. By pre-processing the unstructured documents into a standardized generic format, the system prepares the data in advance for subsequent alignment operations, thereby reducing the time required for alignment to specific use cases while maintaining versatility across multiple ontologies.

Inventive Principle:

Principle #10Preliminary action

3Device complexity

If the system uses a single generic document ontology for all domains, then the system structure is simplified, but the semantic annotations cannot capture domain-specific nuances

Engineering Contradiction:

Improveontology structure simplicityVSAvoiddomain-specific semantic accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the ontology structure into a universal layer (generic document ontology) and domain-specific layers (target ontologies). The generic ontology provides simplified, domain-independent structured annotations, while target ontologies add domain-specific nuances. This segmentation allows the system to maintain structural simplicity at the base level while achieving high precision in domain-specific applications through hierarchical ontology composition.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS12530534B2System and method for generating structured semantic annotations from unstructured document

Publication Date: 2026.01.20 ACCENTURE GLOBAL SOLUTIONS LTD

AI summary

A method and system for generating structured semantic annotations from unstructured documents are disclosed. The method may include obtaining a natural language processing model trained to extract unstructured data from an unstructured document and output structured data based on a generic document ontology. The method may further include inputting a plurality of unstructured documents into the natural language processing model to obtain generic structured annotations of the unstructured documents. The method may further include aligning the generic structured annotations to a target ontology specifying a use case to generate target structured annotations for the use case. The target structured annotations may include entities and relations between the entities. The method may further include generating a knowledge representation for the use case based on the target structured annotations and generic structured annotations.