Hebbian Learning-Based Social Network Recommendations

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing mobile video content recommendation systems fail to effectively utilize user interaction data to provide personalized content recommendations and identify trends within content-based social networks, leading to suboptimal user engagement and content promotion.

Innovation Solution

Implementing a Hebbian-based recommender system that models user interactions as nodes and edges in a network graph, adjusting connection weights based on temporally correlated activities to generate recommendations, identify influencer nodes, and predict trends, thereby enhancing content discovery and user engagement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional content recommendation systems are used, then implementation is simple, but they fail to effectively utilize user interaction data for personalized recommendations and trend identification

Engineering Contradiction:

Improverecommendation accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces traditional rule-based or collaborative filtering recommendation systems with a Hebbian learning-based neural network system. This substitution enables the system to dynamically learn and adapt connection weights between users and content based on interaction patterns, significantly improving recommendation accuracy while accepting increased computational complexity

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system dynamically adjusts connection weights as parameters based on user interaction data. By continuously updating these weights through Hebbian learning (strengthening connections for co-consumed content), the system adapts to changing user preferences and improves recommendation precision over time

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If Hebbian-based learning is implemented to dynamically adjust connection weights, then recommendation accuracy and trend prediction improve, but computational complexity increases

Engineering Contradiction:

Improvepersonalization capabilityVSAvoidcomputational complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The Hebbian learning system operates autonomously by automatically adjusting connection weights based on observed user interactions without requiring manual reconfiguration or complex external optimization algorithms. The system self-adapts to user behavior patterns, providing personalized recommendations while managing computational complexity through localized weight updates rather than global optimization

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system pre-establishes a network graph structure with users, content, and connections before processing interactions. This preliminary structuring allows efficient incremental updates through Hebbian learning rather than requiring complete system reprocessing, reducing computational complexity while maintaining adaptability

Inventive Principle:

Principle #10Preliminary action

3Loss of information

If user interaction data is extensively processed for network modeling, then personalized recommendations and trend identification improve, but data processing time increases

Engineering Contradiction:

Improveuser interaction data utilizationVSAvoiddata processing time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The system extracts only the essential interaction data needed for Hebbian learning (user-content interactions and temporal patterns) from the broader user activity data. By focusing on relevant interaction events rather than processing all user data, the system minimizes data processing time while maintaining effective personalization and trend identification

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system processes user interaction data in periodic batches or streams rather than continuously analyzing every interaction in real-time. This periodic processing approach reduces computational overhead and data processing time while still capturing sufficient interaction patterns for accurate recommendations and trend detection

Inventive Principle:

Principle #19Periodic action

Data Source

PatentUS10827030B2Hebbian learning-based recommendations for social networks

Publication Date: 2020.11.03 VERIZON PATENT & LICENSING INC

AI summary

A network device applies Hebbian-based learning to provide content recommendations in content-based social networks. The method includes obtaining customer activity data for a content-based social network; modeling the customer activity data as nodes and edges within the content-based social network, the nodes representing users and the edges representing connections between the users; assigning initial weights to the edges, that correspond to a connection strength, based on user-designated of relationships between the nodes; adjusting the initial weights in response to temporally correlated activity between the nodes from the customer activity data, to provide adjusted weights; identifying a content recommendation for a particular node based on an activity to access content by another node and one or more of the adjusted weights; storing a customer profile including the content recommendations associated with a node; and providing the content recommendation to a user device associated with the customer profile.