Cluster-Based Random Walking Hash Table Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing random walking algorithms are inefficient for large-scale graph data processing, requiring frequent database access and lacking a scalable solution.
Innovation Solution
A cluster-based method that generates a hash table with node identifiers as keys and adjacent node identifiers as values, allowing for efficient random sequence generation reflecting random walks in graph data, reducing database access and enabling concurrent processing across multiple machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If random walking is implemented by reading nodes and adjacent nodes from a database sequentially, then the algorithm is simple to implement, but the processing efficiency is low and database access overhead is high
Solution Approach 1:
The patent pre-generates hash tables that map node identifiers to their adjacent node identifiers before random walking begins. This preliminary action stores the graph structure in an optimized data structure, eliminating the need for repeated database queries during the random walking process and significantly improving processing efficiency
Solution Approach 2:
The patent creates a copy of the graph data in memory as a hash table structure, rather than repeatedly accessing the original database. This copying approach allows random walking to operate on the in-memory representation, reducing database access overhead while maintaining the graph's structural information
2Adaptability or versatility
If graph data is processed using traditional random walking methods, then the implementation is straightforward, but it cannot scale to large-scale graph data efficiently
Solution Approach 1:
The patent transitions from a one-dimensional sequential database access model to a two-dimensional structure by creating hash tables that provide both node identification and adjacent node mapping in a single data structure. This dimensional change enables O(1) average time complexity for accessing adjacent nodes, making the algorithm scalable to large-scale graph data
Solution Approach 2:
The patent changes the data access parameter from sequential database queries to hash table lookups. By transforming the data storage and access mechanism, the algorithm achieves constant-time access to adjacent nodes regardless of graph size, enabling efficient processing of large-scale graphs
3Speed
If a hash table is generated to store node and adjacent node relationships, then random sequence generation becomes efficient, but additional memory consumption increases
Solution Approach 1:
Instead of storing all graph data in the traditional database format and querying it during random walking, the patent inverts the approach by pre-computing and storing only the essential adjacency relationships in hash tables. This inversion reduces the data footprint while maintaining fast access capabilities
Data Source
AI summary
Embodiments of the present specification disclose random walking and a cluster-based random walking method, apparatus and device. A solution includes: obtaining information about each node included in graph data, generating, according to the information about each node, a hash table reflecting a correspondence between the node and an adjacent node of the node, and generating a random sequence according to the hash table, to implement random walking in the graph data. The solution is applicable to clusters and single machines.