NLP Feature Extraction for Alias-Aware Watchlist Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional watchlist identification methods rely heavily on manual review of structured data sources, leading to errors due to name misspellings, incorrect PII, and lack of consideration for aliases, while external unstructured data sources like social media and news articles are underutilized due to difficulty in normalization and relevance assessment.
Innovation Solution
A system and method utilizing natural language processing and machine learning to process and normalize external data sources, constructing holistic identity profiles through unsupervised semantic identity modeling and graph-based clustering, enabling accurate watchlist matching and transaction risk assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If manual review of traditional watchlist data is used, then simplicity of the system is maintained, but identification accuracy deteriorates due to name misspellings, incorrect PII, and lack of alias consideration
Solution Approach 1:
The patent replaces manual mechanical review processes with automated natural language processing and machine learning systems. The NLP engine automatically processes watchlist data, performs name normalization, identifies aliases, and extracts PII without human intervention, thereby maintaining system simplicity while dramatically improving identification accuracy through intelligent algorithms.
Solution Approach 2:
The system changes the parameters of data processing by transforming unstructured text data into structured, normalized formats. It applies parameter transformations including name standardization, alias expansion, and PII extraction, converting raw watchlist entries into comparable, machine-processable formats that enhance matching accuracy without requiring complex manual procedures.
2Loss of information
If external unstructured data sources are utilized, then data completeness is improved, but data processing complexity increases due to normalization and relevance assessment difficulties
Solution Approach 1:
The patent implements a universal NLP processing pipeline that handles multiple types of external data sources (social media, news articles, web content) through a single integrated system. This multi-functional engine performs extraction, normalization, and relevance assessment across diverse formats, reducing processing complexity by applying consistent methods universally rather than requiring separate handling for each data type.
Solution Approach 2:
The system introduces an intermediary NLP layer between raw external data sources and the watchlist matching system. This intermediary performs text extraction, entity recognition, and normalization, converting unstructured external data into a standardized format that the core matching system can process efficiently, thereby reducing overall processing complexity.
3Ease of operation
If traditional structured data sources are used, then data processing ease is maintained, but decision accuracy deteriorates due to limited information availability
Solution Approach 1:
The system performs preliminary processing of external unstructured data sources before they reach the decision-making stage. The NLP engine pre-extracts relevant information, normalizes formats, and flags potential matches in advance, so that when structured watchlist data is compared, the system already has enriched contextual information available, improving decision accuracy without complicating the core operation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a method and system for identity correlation between a transaction applicant (TA) and a watchlist entity (WE). Preexisting watchlist data and other aggregated identity data (AID) are processed to provide for comparison to a collective identity of at least the TA. Using various categorizations for the AID and the collective identity, watchlist tags are generated that can then be matched to the collective identity. As a result of the matching, a watchlist candidacy demonstrating a probability that the identity of the TA does or does not correspond to that of the WE can be generated. Natural language processing may be used in connection with sourced textual information such as from news articles, social media posts and the like to supplement the AID available to the system and to further enhance matching functionality.