Parallel Top-K Shortest Path Discovery in Distributed Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing query techniques face challenges in efficiently determining top-K simple shortest paths in large-scale graph data, particularly in distributed networks, where scalability and loop detection are critical for applications like social networks and web graph analysis.
Innovation Solution
A distributed method across multiple computing servers, where graph data is partitioned and processed in parallel, with asynchronous message exchange and adaptive cutoff thresholds to filter out loops and determine top-K shortest paths, ensuring efficient convergence and accurate results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If graph data is processed in a distributed network with multiple computing servers, then scalability is improved, but system complexity increases
Solution Approach 1:
The patent divides the graph data into partitions distributed across multiple computing servers. Each server independently processes its local partition, enabling parallel computation and scaling to large datasets without requiring a single centralized system to handle all data.
Solution Approach 2:
The patent introduces a coordinator node that acts as an intermediary to manage the distributed computation. The coordinator collects results from various computing servers, performs global ranking to determine top-K shortest paths, and coordinates the overall algorithm execution, thereby simplifying the complexity of direct peer-to-peer communication between all servers.
2Productivity
If parallel processing is used across multiple computing servers, then query performance is improved, but communication overhead increases
Solution Approach 1:
The patent performs preliminary local processing at each computing server by computing shortest paths within local graph partitions before exchanging results. This preliminary action reduces the volume of data that needs to be communicated between servers, as only relevant path information rather than complete graph data is exchanged.
Solution Approach 2:
The patent extracts only the necessary path information from each computing server's local processing results and sends only this extracted data to the coordinator. This selective extraction minimizes communication overhead by transmitting only the essential information needed for global top-K ranking rather than complete path details.
3Measurement precision
If loop detection is implemented in path sequences, then result accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs loop detection as a preliminary filtering step during the path sequence generation process. By checking for loops early in the computation rather than after complete path enumeration, the system ensures accuracy while minimizing the time penalty, as loop-free paths are identified and discarded before extensive processing occurs.
Solution Approach 2:
The algorithm inherently detects loops through the nature of shortest path computation in distributed systems. Each computing server independently identifies loops in its local path sequences using standard graph traversal techniques, and the coordinator aggregates these loop-free results. The system's structure naturally facilitates loop detection without requiring additional complex mechanisms.
Data Source
AI summary
A method for searching the top-K simple shortest paths between a specified source node and a specified target node in a graph, with graph data partitioned and distributed across a plurality of computing servers, the method including a parallel path search initialized from either one or both of the source and target nodes and traversing the graph by building likely path sequences for a match. Each computing server determines and forwards a path sequence as discovery progresses until the top-K paths are discovered.