Merchant Category Code Classification via Social Media NLP

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Smaller businesses face challenges in selecting the appropriate merchant category code due to the large number of available codes and may incorrectly categorize themselves to avoid regulations, leading to difficulties in obtaining merchant accounts for credit card payments.

Innovation Solution

A method and system that utilize natural language processing on social media information to identify key terms and calculate frequency values to determine the matching merchant category code, integrating with a machine learning framework to predict the correct merchant category code.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a merchant manually selects a merchant category code from over seven hundred codes, then the merchant can apply for a merchant account, but the complexity of selection increases and accuracy decreases

Engineering Contradiction:

Improveease of merchant category code selectionVSAvoidaccuracy of merchant category code classification

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The system performs automatic merchant category code classification using machine learning models that analyze merchant information from multiple sources (administrative information, social media data, website content) to autonomously determine the appropriate category code without requiring manual merchant selection, thereby simplifying the process while improving accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent introduces an intermediary classification system that acts as a mediator between the merchant and the merchant account application process. This intermediary automatically analyzes merchant data and recommends or assigns category codes, reducing the burden on merchants while ensuring accurate classification

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If merchants are provided with detailed information about all merchant category codes, then accurate classification can be achieved, but the complexity of the application process increases

Engineering Contradiction:

Improveaccuracy of merchant category code classificationVSAvoidcomplexity of classification system

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The classification system is segmented into multiple independent components: data collection modules that gather information from various sources, natural language processing modules that analyze text data, machine learning models that perform classification, and validation modules that verify results. This segmentation allows each component to specialize in specific tasks, improving overall accuracy while managing complexity through modular design

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent creates a universal classification framework that handles multiple data types (structured administrative information, unstructured social media posts, website content) and various merchant categories through a single multi-functional system. This universal approach improves accuracy by considering multiple information sources while managing complexity through a unified architecture

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Ease of operation

If merchants select category codes to avoid regulations, then ease of operation is improved, but reliability of classification deteriorates

Engineering Contradiction:

Improveease of merchant account applicationVSAvoidreliability of merchant category code assignment

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The system implements feedback mechanisms where classification results are validated against multiple data sources and regulatory requirements. The machine learning models are trained on labeled data that includes correct category assignments, providing feedback that guides the classification process toward reliable and compliant results while maintaining ease of operation through automation

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent applies preliminary anti-action by proactively preventing incorrect or regulatory-violating category code selections before they can be submitted. The system analyzes merchant information and predicts appropriate category codes, blocking or correcting potential misclassifications that would violate regulations, thereby maintaining both ease of operation and reliability

Inventive Principle:

Principle #9Preliminary anti-action

Data Source

PatentUS11107139B1Computing system learning of a merchant category code

Publication Date: 2021.08.31 INTUIT INC

AI summary

A method for computing system learning of a merchant category code includes accessing a social media platform using administrative information of a merchant to obtain social media information of the merchant, and performing natural language processing on the social media information to add key terms to a merchant terminology corpus defined for the merchant. The key terms in the first merchant terminology corpus are normalized. For each key term in the first merchant terminology corpus and for each merchant category code, a frequency value for the merchant category code is calculated to obtain multiple frequency values. The method further includes aggregating frequency values across the first key terms in the first merchant terminology corpus to select the matching merchant category code of the merchant, and applying the matching merchant category code to the merchant.