Geographic Information Validation Using OCR and Heuristic Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for validating and normalizing geographic information in business transactions, especially across international standards, are inefficient due to varying address formats and the lack of accessible reference data, leading to challenges in automatic extraction and recognition of relevant information.
Innovation Solution
A computer program product that performs optical character recognition (OCR) on document images, extracts identifiers, and compares them with global address information from proprietary and open-source databases using predefined heuristic rules to normalize and validate address information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If proprietary address database services are used for validation, then validation accuracy is improved, but cost increases significantly
Solution Approach 1:
The patent creates a copy of address validation functionality by implementing a hybrid database system that replicates the capabilities of proprietary services. It combines open-source address data with machine learning models to produce a local copy of validation functionality, eliminating the need for expensive continuous subscriptions to external proprietary services while maintaining validation accuracy.
Solution Approach 2:
The system uses inexpensive open-source address databases as temporary, disposable data sources that can be freely updated and replaced. Rather than investing in expensive long-term proprietary services, the patent leverages freely available data that can be quickly refreshed from multiple open-source providers, reducing overall cost while maintaining data freshness.
2Adaptability or versatility
If multiple address database sources are integrated, then global coverage is improved, but system complexity increases
Solution Approach 1:
The patent introduces a machine learning model as an intermediary layer between multiple open-source address databases and the validation interface. This mediator automatically harmonizes data from diverse sources, resolving conflicts and standardizing formats without requiring complex manual integration logic. The ML model acts as a smart intermediary that simplifies the complexity of managing multiple data sources.
Solution Approach 2:
The system merges multiple open-source address databases into a unified hybrid validation system. By combining data from various free sources with machine learning capabilities, it creates a consolidated solution that achieves global coverage equivalent to proprietary services but with reduced operational complexity through automated data harmonization.
3Reliability
If manual review of document content is performed, then validation accuracy is improved, but time consumption increases
Solution Approach 1:
The patent implements self-service validation where the machine learning model automatically performs the verification task that would otherwise require manual review. The system extracts address information from documents and validates it against the hybrid database autonomously, providing accurate validation results without human intervention and eliminating the time cost of manual review while maintaining high accuracy.
4Reliability
If proprietary address services are subscribed to, then data authority is improved, but accessibility decreases
Solution Approach 1:
The patent creates a locally accessible copy of authoritative address validation functionality by training machine learning models on proprietary data and implementing a hybrid database system. This local copy provides the same data authority as proprietary services but with unrestricted accessibility, allowing the organization to perform validations independently without relying on external service availability or subscription status.
Data Source
AI summary
A computer program product includes program instructions configured to cause a processor, to: perform optical character recognition (OCR) on an image of a document; extract an identifier of the document from the image based at least in part on the OCR; compare at least portions of the identifier with content from one or more reference data sources; and determine whether the identifier is valid based at least in part on the comparison. The content comprises global address information; while the content from the reference is derived from geographic information. Deriving the content from the geographic information includes: obtaining the geographic information; and parsing the geographic information according to a set of predefined heuristic rules, where the heuristic rules are configured to normalize the global address information obtained from the one or more sources according to a single convention for representing address information.