P2P Transaction Text Analysis for Fraud Indicator Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

P2P transaction platforms face challenges in detecting fraudulent or illicit activities due to anonymity, high transaction volumes, evolving fraud tactics, and complex transaction patterns, leading to resource-intensive detection and false positives.

Innovation Solution

Implementing natural language processing (NLP) and machine learning techniques to analyze textual information in P2P transactions, identifying indicators of fraud or illicit activity through supervised learning, ensemble methods, and NER, to trigger remediation actions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional detection methods are used for P2P transactions, then resource consumption is high and false positives increase, but detection accuracy remains insufficient

Engineering Contradiction:

Improvedetection accuracyVSAvoidresource consumption

Core Design Contradiction:

Measurement precisionVSLoss of energy

Solution Approach 1:

The system segments fraud detection into multiple specialized components: NLP model for textual analysis, machine learning model for pattern recognition, and rule-based systems for specific fraud indicators. Each component processes specific aspects of transaction data independently, improving detection accuracy while distributing computational load efficiently

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary NLP processing layer that translates unstructured textual information from memos and descriptions into structured features. This intermediary layer enables more efficient downstream processing by converting raw text into meaningful indicators that can be quickly evaluated by detection algorithms

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If comprehensive text analysis is performed on all P2P transactions, then detection accuracy improves, but processing time increases

Engineering Contradiction:

Improvedetection accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system applies partial text analysis by focusing computational resources on specific high-risk indicators within transaction text rather than analyzing all text uniformly. The NLP model identifies and prioritizes certain linguistic patterns and phrases that are more indicative of fraud, performing detailed analysis only on relevant portions of the text

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent changes the parameter of text analysis from comprehensive to selective by adjusting the sensitivity and scope of NLP processing based on transaction risk indicators. The system dynamically adjusts which textual features are extracted and analyzed based on initial risk assessment signals from other data points

Inventive Principle:

Principle #35Parameter changes

3Reliability

If manual review processes are used for fraud detection, then false positives can be reduced, but productivity decreases

Engineering Contradiction:

Improvefalse positive reductionVSAvoidtransaction processing throughput

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system implements self-service fraud detection through automated NLP and machine learning models that independently analyze transaction text and identify fraud indicators without requiring manual review for every transaction. The system serves itself by automatically flagging suspicious transactions and only escalating cases that exceed configured risk thresholds

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent incorporates feedback mechanisms where detection results and outcomes are used to continuously refine and improve the NLP and machine learning models. The system learns from confirmed fraud cases and false positives, adjusting its detection parameters and text analysis focus to reduce false positives while maintaining high throughput

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20260037975A1Detecting fraudulent or illicit activity in peer-to-peer transactions using natural language processing

Publication Date: 2026.02.05 CAPITAL ONE SERVICES LLC

AI summary

In some implementations, a system may obtain information associated with indicators related to fraudulent or illicit activity in P2P transactions. The system may receive, from a first user device, a request for a P2P transaction. The system may analyze textual information related to the P2P transaction using natural language processing (NLP) to determine whether the P2P transaction includes one or more of the indicators related to fraudulent or illicit activity. The system may process the request for the P2P transaction in accordance with whether the P2P transaction includes one or more of the indicators related to fraudulent or illicit activity. For example, the system may trigger a remediation action based on the textual information including indicators related to fraudulent or illicit activity, or may process the P2P transaction based on the textual information lacking indicators related to fraudulent or illicit activity or including indicators of legitimate activity.