Name Matching Hybrid System for Identity Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Document management systems face challenges in accurately verifying identities and matching names, particularly when variations in naming conventions occur, leading to issues with granting access to electronic documents.
Innovation Solution
A hybrid approach combining machine learning technology and rule-based methods for name matching, utilizing a database of matched name pairs, phonetic indexing, similarity distance metrics, and name matching rules to allow for variations such as case insensitivity, transliterations, and diacritical conversions, enhancing the matching process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exact name matching is used, then identity verification accuracy is improved, but access is denied to users with variations in naming conventions
Solution Approach 1:
The system changes the matching parameters from exact string equality to semantic similarity based on phonetic indexing and machine learning models. This allows the system to accept names that are phonetically similar or semantically equivalent, resolving the contradiction between exact matching accuracy and adaptability to name variations.
Solution Approach 2:
The patent replaces the mechanical exact-string-matching system with a machine learning-based semantic matching system. The machine learning model processes name strings, applies phonetic indexing, and determines similarity based on learned patterns, substituting rigid mechanical matching with adaptive intelligent processing.
2Reliability
If hybrid name matching techniques are applied, then name matching success rate is improved, but system complexity increases
Solution Approach 1:
The hybrid matching system is segmented into distinct components: phonetic indexing module, machine learning similarity module, and rule-based verification module. Each component handles specific aspects of name matching, allowing the complex system to be modular and maintainable while achieving high reliability through coordinated operation of these segments.
Solution Approach 2:
The patent introduces phonetic indexing as an intermediary representation between the raw name strings and the machine learning model. This intermediary layer simplifies the comparison process by transforming names into phonetic codes, reducing the complexity of direct string comparison while maintaining high matching accuracy.
3Measurement precision
If machine learning approach is used, then recognition of similar names is improved, but computational resources are consumed
Solution Approach 1:
The system performs preliminary phonetic indexing of names before they are processed by the machine learning model. This preliminary action pre-processes the data into a format that is more suitable for efficient machine learning comparison, reducing the computational burden during the actual similarity assessment phase.
Solution Approach 2:
The machine learning model applies different processing depths to different aspects of name comparison. Rather than uniformly processing all name characteristics, the model focuses computational resources on the most discriminative features for name similarity, optimizing the balance between recognition accuracy and resource consumption.
Data Source
AI summary
Techniques are described for recognizing alternative representations of the same name. An example method includes transmitting, by a document management platform implemented by a computing system, a document package to a second computing device. The document package includes a document received from a first computing device and an indication of first name. The document management platform obtains an indication of a second name from an identity document provided by a user of the second computing device. The document management platform performs a name matching operation using a machine learning model to determine whether the first name and the second name are similar based on a similarity score generated by the machine learning model. Based on determining that the first name and the second name are similar, the document management platform grants the user of the second computing device access to the document.