Graph-Based Publish-Subscribe Matching for Semantic Subscriptions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current publish/subscribe communication tools lack an efficient mechanism to manage dynamic graph-based subscriptions and publications, limiting their ability to express consumer interest in semantic information domains effectively.
Innovation Solution
A graph-based publish/subscribe method is introduced, where a graph with nodes and edges represents the information domain, allowing publishers to publish content on selected nodes and edges, and subscribers to subscribe to specific subsets, with attributes and queries used to determine matching interests, and dynamic modifications to the graph are handled to maintain relevant subscriptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional topic-based subscription language is used, then implementation is simple, but the ability to express consumer interest in semantic information domains is limited
Solution Approach 1:
The patent transforms the subscription language from traditional topic-based parameters to graph-based parameters (nodes, edges, paths, queries). This parameter change enables subscribers to express complex semantic interests using graph query languages while maintaining a structured and manageable subscription mechanism.
Solution Approach 2:
The patent introduces a graph-based dimension to the subscription language, moving from flat topic names to multi-dimensional graph structures with nodes, edges, and paths. This dimensional enhancement allows subscribers to express complex semantic relationships and interests that cannot be captured by simple topic strings.
2Adaptability or versatility
If content-based PUBSUB systems are used with query languages, then subscription expressiveness improves, but matching efficiency decreases
Solution Approach 1:
The patent performs preliminary actions by pre-processing subscription queries into graph-based representations and pre-establishing matching rules. When publications arrive, the system can quickly match them against pre-processed subscription patterns, significantly improving matching efficiency while maintaining high expressiveness.
Solution Approach 2:
The patent creates simplified copies or representations of complex graph queries that can be efficiently matched against publications. By transforming complex semantic queries into standardized graph patterns, the system maintains expressiveness while enabling faster matching operations.
3Adaptability or versatility
If dynamic graph modifications are supported, then system adaptability improves, but system complexity increases
Solution Approach 1:
The patent implements dynamic graph modifications that allow the information domain model to adapt in real-time. Nodes, edges, and queries can be added or modified dynamically, enabling the system to adapt to changing semantic domains while maintaining a relatively simple underlying graph structure.
Solution Approach 2:
The patent enables the graph-based subscription system to automatically handle dynamic modifications through self-service mechanisms. When the graph structure changes, the system automatically updates subscription matchings and maintains consistency without requiring complex manual intervention.
4Productivity
If graph-based subscriptions are used, then the number of matching subscriptions decreases, but broker traffic increases
Solution Approach 1:
The patent extracts and filters out non-matching subscriptions early in the processing pipeline using graph-based pattern matching. By identifying and removing non-matching subscriptions before they reach the broker, the system reduces the number of subscriptions that need to be processed, thereby reducing broker traffic despite the increased complexity of graph-based matching.
Data Source
AI summary
A system and method for graph-based publication/subscription are provided. A graph comprising nodes and edges is created, each node representative of a point of interest in an information domain, each edge linking a first node and a second node and representative of a relationship between a first and a second point of interest. A request to publish content on at least one selected node and/or at least one selected edge is received. A request to subscribe to a subset of nodes and/or a subset of edges is received from a subscriber. The subscription request is correlated with the publication request to determine whether the subset of nodes comprises the selected node and whether the subset of edges comprises the selected edge. The content is notified to the subscriber if the subset of nodes comprises the selected node and the subset of edges comprises the selected edge.