Product Title Quantity Extraction for Similarity Comparison

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional computer systems are limited in accurately determining product similarity from online product listings due to the lack of universal product identifiers, leading to inaccurate comparisons and a lack of comprehensive understanding of competing products.

Innovation Solution

A computer-implemented system that extracts quantities from product titles by analyzing tags, determining quantities based on category and unit, and generating similarity values between product identifiers, enabling accurate comparisons and similarity determinations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional computer systems use product identifiers (UPC, EAN) to determine product similarity, then product identification accuracy is improved, but the system becomes useless in areas where these identifiers are not widely used

Engineering Contradiction:

Improveproduct identification accuracyVSAvoidglobal applicability

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent introduces natural language processing as an intermediary mechanism that bridges the gap between product listings and computer systems. Instead of relying solely on standardized identifiers, the system uses NLP to extract and compare product attributes from text descriptions, enabling accurate product similarity determination across diverse global markets where universal identifiers are not universally adopted.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system achieves universality by combining multiple approaches: traditional product identifier parsing and natural language processing. This multi-functional approach allows the same system to accurately process products with standardized identifiers (UPC, EAN) while simultaneously handling products described through natural language, making it globally applicable across different regions and markets.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Productivity

If computer systems compare product listings using conventional methods, then processing speed is maintained, but accuracy of similarity determination deteriorates due to inability to understand different words and descriptors

Engineering Contradiction:

Improveprocessing speedVSAvoidsimilarity determination accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary processing by extracting and normalizing product attributes from product titles and descriptions before comparison. Natural language processing techniques pre-process the text data to identify and standardize product characteristics, enabling accurate similarity determination while maintaining efficient processing speeds through structured data representation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent transforms product description parameters from unstructured text to structured normalized attributes. By converting various product descriptors into standardized parameter formats, the system enables accurate comparison while maintaining processing efficiency. The parameter transformation allows different wordings to be mapped to equivalent structured representations for reliable similarity determination.

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If humans manually compare product listings to determine similarity, then comprehensive understanding of products is achieved, but time consumption increases significantly

Engineering Contradiction:

Improvecomprehensive product understandingVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs self-service by automatically extracting, processing, and comparing product attributes using natural language processing algorithms. Instead of requiring human intervention for product similarity determination, the system independently processes product listings, extracts relevant features, and generates similarity assessments, dramatically reducing time consumption while maintaining comprehensive understanding.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical manual comparison process with automated natural language processing mechanisms. The system uses computational algorithms to analyze product descriptions, extract attributes, and determine similarity, substituting human cognitive processing with automated digital processing that achieves the same comprehensive understanding much faster.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS11798056B2Systems and methods for intelligent extraction of quantities from product titles

Publication Date: 2023.10.24 COUPANG CORP

AI summary

Some aspects of the present disclosure are directed to computerized methods for extracting attributes from product titles. The method may include: retrieving first product identifier comprising at least one tag; determining, based on the at least one tag, a number of quantity related tags; flagging the product identifier as having a quantity based on an analysis of the quantity related tags; comparing the at least one tag and quantity of the first product identifier with at least one tag and quantity associated with a second product identifier; generating, based on the comparison, at least one similarity value between the first product identifier and the second product identifier; and transmitting instructions to at least one user device, wherein the instructions cause the at least one user device to display the at least one similarity value.