Automated Item Compatibility Inference via Knowledge Graph Embeddings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for determining item compatibility rely heavily on manually curated databases, which are often outdated and prone to errors, leading to missed sales, increased returns, and reduced product visibility, especially when shoppers require expertise to understand technical attributes.
Innovation Solution
An item compatibility system that automatically infers compatibility by extracting information from structured and unstructured data using natural language processing and knowledge graphs, generating embeddings to identify compatible items and attributes, and providing a customized user experience through a prediction component and generative model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manually curated compatibility databases are used, then compatibility information can be provided, but the system becomes outdated and error-prone, leading to missed sales and increased returns
Solution Approach 1:
The system automatically extracts compatibility information from product listings and updates the knowledge graph without manual intervention. The NLP component processes unstructured data from product descriptions, and the embedding component automatically infers compatibility relationships, enabling the system to self-update and maintain current compatibility information across the marketplace.
Solution Approach 2:
The patent replaces manual database curation with automated NLP-based information extraction and machine learning models. Instead of manually maintaining structured compatibility databases, the system uses neural networks to process unstructured product data and automatically infer compatibility relationships, substituting mechanical manual processes with automated intelligent systems.
2Adaptability or versatility
If structured compatibility databases are used, then compatibility can be determined, but new items and incorrect information are not addressed, reducing product visibility
Solution Approach 1:
The system performs preliminary extraction of compatibility information from product listings as they are added to the marketplace. The NLP component processes unstructured data from new product descriptions and the embedding component infers compatibility relationships before the items become prominent, ensuring new products are properly catalogued and visible from the start.
Solution Approach 2:
The system continuously monitors and updates compatibility information by processing new product listings and feedback from the marketplace. The NLP component analyzes unstructured data from product descriptions and reviews, and the embedding component adjusts compatibility relationships based on observed usage patterns and user interactions, creating a feedback loop that maintains accurate and current compatibility information.
3Measurement precision
If expertise is required to understand technical attributes, then compatibility can be accurately determined, but the search process becomes time-consuming and complex
Solution Approach 1:
The patent introduces an intermediary NLP-based compatibility inference system that bridges the gap between users and complex technical compatibility information. Instead of requiring users to manually interpret technical specifications, the system automatically processes product data, extracts compatibility attributes, and presents simplified compatibility information, acting as an intelligent mediator that translates complex data into useful insights.
Data Source
AI summary
Systems and methods for inferring compatibility relationships are described. Embodiments of the present disclosure identify user interaction history including an interaction between a user and a first product, wherein the first product comprises an attribute that is compatible with a subset of available products; query a database that includes the available products to identify a second product from the subset of available products based on the attribute, wherein the second product is identified based on a knowledge graph that includes a first node representing the first product and a second node representing the second product; and provide a customized user experience for the user that indicates the second product and the attribute.