Edge Label Propagation for Network Relationship Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for inferring relationships in online networks lack accuracy in accounting for topological features and homophily, leading to inefficient content retrieval and user interaction within the network.
Innovation Solution
The method involves training a machine learning model using labeled edges to predict relationship attributes, which are then propagated across the network, combining edge scores and topological features to infer likelihoods of relationships for unlabeled edges, with a near-linear time complexity for efficient propagation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing methods are used to infer relationships in online networks, then the process is simpler, but the accuracy in accounting for topological features and homophily is insufficient
Solution Approach 1:
The patent introduces label propagation as an intermediary mechanism that mediates between the network structure (topological features) and the relationship inference process. The label propagation algorithm acts as a mediator that systematically accounts for homophily by propagating relationship labels through the network structure, thereby improving measurement precision without requiring direct complex calculations for each node pair.
Solution Approach 2:
The patent replaces traditional mechanical or direct computational methods of relationship inference with a machine learning model that incorporates topological features and homophily. This substitution allows the system to automatically learn and account for complex network patterns, improving accuracy while the system manages the computational complexity through algorithmic efficiency.
2Measurement precision
If complex inference methods are used to account for topological features and homophily, then the accuracy improves, but the computational efficiency decreases
Solution Approach 1:
The patent changes the parameters of the inference process by incorporating topological features and homophily into the machine learning model. By adjusting these parameters and using label propagation, the system achieves higher accuracy while maintaining computational efficiency through the iterative nature of the propagation algorithm, which converges relatively quickly on large networks.
3Measurement precision
If label propagation is performed across the entire network, then the coverage is complete, but the time complexity increases
Solution Approach 1:
The patent segments the label propagation process by focusing on local neighborhoods and using iterative updates that propagate labels progressively through the network. This segmentation allows the system to achieve complete coverage without processing the entire network at once, thereby reducing time complexity while maintaining completeness of inference.
Data Source
AI summary
The disclosed embodiments provide a system for performing inference. During operation, the system obtains a graph containing nodes representing members of an online system, edges between pairs of nodes, and edge scores representing confidences in a type of relationship between the pairs of nodes. Next, the system performs a set of iterations that propagate a label for the type of relationship from a first subset of edges to remaining edges in the graph, with each iteration updating a probability of the label for an edge between a pair of nodes based on a subset of edge scores for a second subset of edges connected to one or both nodes in the pair and probabilities of the label for the second subset of edges. The system then performs one or more tasks in the online system based on the probability of the label for the edge.