ML Transaction Supervision Event Detection With Rule-Based Screening

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing financial transaction supervision systems lack effective methods to identify and manage events of interest using machine learning, leading to inefficiencies in transaction analysis and oversight.

Innovation Solution

A system utilizing machine learning models to classify transactions, generate attributes, apply predefined rules, and evaluate criteria to identify supervisory events, with optional features for generating alerts and recommended actions based on transaction characteristics and historical data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If machine learning models are utilized to classify financial transactions and generate attributes, then the precision of transaction analysis is improved, but the complexity of the supervision system increases

Engineering Contradiction:

Improvetransaction analysis precisionVSAvoidsupervision system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the complex supervision task into distinct components: machine learning models for classification and attribute generation, rule-based systems for event identification, and separate processing stages. This modular segmentation allows each component to specialize in specific functions, improving overall analysis precision while managing system complexity through structured division of labor.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary layer that bridges machine learning outputs and rule-based processing. Generated transaction attributes serve as intermediaries that translate complex ML model outputs into structured data that can be efficiently processed by rule-based systems, thereby maintaining precision while reducing the computational burden on the overall system.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If multiple machine learning models are trained on historical transactions to generate transaction attributes, then the reliability of event identification is improved, but the time required for processing increases

Engineering Contradiction:

Improveevent identification reliabilityVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-training multiple machine learning models on historical transaction data before actual supervision tasks. These models generate transaction attributes in advance that can be stored and reused, allowing the system to maintain high reliability in event identification while reducing processing time during actual transaction supervision through cached attribute data.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent applies partial action by selectively using only the necessary subset of generated attributes and models for each specific transaction type or supervision scenario. Rather than processing all possible attributes from all models for every transaction, the system identifies and processes only the relevant subset, maintaining reliability where needed while significantly reducing overall processing time.

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If transaction matching criteria are evaluated for each classified transaction to form selected transactions, then the accuracy of supervisory event detection is improved, but the productivity of the system decreases

Engineering Contradiction:

Improveevent detection accuracyVSAvoidtransaction processing productivity

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The system applies local quality by tailoring the evaluation of matching criteria to specific transaction types, risk categories, or supervisory contexts. Rather than uniformly applying complex matching criteria to all transactions, the system adjusts the depth and specificity of criterion evaluation based on local characteristics of each transaction or category, maintaining high detection accuracy for critical transactions while improving overall productivity through differentiated processing.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20260080345A1Machine-Learning-Based Identification of Financial Transaction Supervision Events

Publication Date: 2026.03.19 STATEMENT TECHNOLOGIES LTD

AI summary

There is provided a processor-based system of identifying a financial transactions supervisory event, the processor configured to: utilize machine learning models to classify a plurality of financial transactions of an organization, thereby resulting, for each of the financials transactions, in one or more respective generated transaction attributes, apply a predefined transaction rule to the classified transactions, the applying comprising: evaluating a transaction matching criterion that is at least partially based on one or more of the respective generated attributes of the classified transactions, the classified transactions matching the transaction matching criterion thereby constituting a set of selected transactions, determining a selected transactions characteristic (STC), based on, one or more transaction characteristics of the transactions constituted in the set of selected transactions, evaluating a supervisory event criterion (SEC), the supervisory action criterion being based on the STC, and identifying a financial transactions supervisory event, responsive to positive evaluation of the SEC.