Graph Neural Recommendation Using Self-Attention Attribute Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Consumers face difficulty in finding products they want to purchase due to the vast array of options available in electronic commerce, necessitating improved item recommendation technologies.
Innovation Solution
A computer-implemented method utilizing a graph neural network model constructed through vectorized representations of entity attributes, including self-attention and neighborhood aggregate operations, to enhance item recommendation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a large number of products are made available in electronic commerce, then product variety and selection are improved, but it becomes difficult for consumers to find the products they want to purchase
Solution Approach 1:
The patent replaces manual product searching and browsing (mechanical system) with an automated recommendation system based on graph neural networks and attention mechanisms. The system automatically analyzes product attributes, constructs knowledge graphs, and generates personalized recommendations, substituting the manual search process with intelligent algorithms that calculate user preference probabilities and rank products accordingly.
2Measurement precision
If traditional recommendation methods are used, then implementation simplicity is maintained, but recommendation accuracy and user preference calculation precision are insufficient
Solution Approach 1:
The patent segments the recommendation system into distinct functional modules: attribute extraction module, knowledge graph construction module, graph neural network processing module, and recommendation generation module. Each module performs a specific function - extracting product attributes, constructing entity relationships, processing embeddings with attention mechanisms, and generating final recommendations. This segmentation allows the complex system to be managed through modular components while achieving high accuracy.
Solution Approach 2:
The patent introduces vector embeddings as an intermediary representation between raw product attributes and recommendation decisions. Product attributes are transformed into dense vector representations, which are then processed by graph neural networks and attention mechanisms. This intermediary layer enables sophisticated pattern recognition and preference calculation while maintaining system manageability through standardized transformation processes.
Data Source
AI summary
A computer-implemented method is provided. The computer-implemented method includes obtaining one or more attributes of one or more entities; performing a vectorized representation on a respective attribute of the one or more attributes, the vectorized representation being encoded as a vector of d dimensions; and performing computation on vectorized representation of the one or more attributes based on attention mechanism to construct a graph neural network model. Vectors of m number of dimensions are used to represent an individual entity in the graph neural network model, (m−1) dimensions out of m dimensions represent (m−1) number of attributes, and one out of the m dimensions represents a flag bit. Constructing the graph neural network model includes performing self-attention on the individual entity, during which the flag bit is converted into a fused flag bit by fusing the individual entity's own attributes.