Distributed Key-Value Store for Real-Time Social Network Path Traversal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graph-oriented databases, such as Pregel and Neo4J, face limitations in real-time search capabilities and horizontal scalability, making them inadequate for efficiently managing and operating large social networks.
Innovation Solution
A distributed memory system using a key/value schema stores social networks as graphs, allowing for fast and efficient set operations like intersection and union, enabling real-time inference of relationships by traversing the graph and reconstructing paths between nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If traditional graph-oriented databases (Pregel, Neo4J) are used to store and process social networks, then graph data structure support is provided, but real-time search capability and horizontal scalability are limited
Solution Approach 1:
The patent segments the graph database into two separate components: a distributed key-value store for node data storage and a separate processing engine for graph operations. This segmentation allows each component to be independently optimized and scaled horizontally, resolving the contradiction between maintaining graph structure support and achieving real-time search capability with horizontal scalability.
Solution Approach 2:
The patent introduces an intermediary layer that translates graph-oriented queries into key-value store operations. This intermediary processing engine acts as a mediator between the user's graph queries and the underlying distributed key-value storage, enabling real-time search capabilities while maintaining compatibility with horizontal scalability of the storage layer.
2Productivity
If graph-oriented databases are used to manage large social networks, then graph structure representation is maintained, but real-time query performance and scalability are insufficient
Solution Approach 1:
The patent replaces the traditional mechanical graph traversal mechanism with a set-based mathematical approach using distributed key-value operations. By substituting iterative graph algorithms with parallel set operations (union, intersection, difference) on neighbor sets stored in the distributed key-value store, the system achieves significantly improved query operation efficiency and reduced processing time for real-time social network analytics.
3Speed
If distributed memory with key/value schema is used, then data access speed and horizontal scalability are improved, but native graph structure support is lost
Solution Approach 1:
The patent makes the distributed key-value store universal by designing it to handle both simple key-value storage and complex graph operations. The system stores graph nodes and their neighbor sets in the key-value store, and the processing engine can perform various graph operations (path finding, community detection, influence analysis) using set operations on this stored data, thus providing multi-functionality that bridges the gap between distributed storage efficiency and graph structure support.
Data Source
AI summary
Systems and methods for managing and evaluating a social network. The social network is represented as a graph structure and stored in distributed memory. A viable path from one node that is not directly connected to another node in the graph structure may be determined by traversing the graph in stages, moving outward from each node in stages until common midpoint nodes are found providing a connection between the nodes. When midpoint nodes are found, the paths connecting the one node to the other node may be reconstructed.