Relation-Tree Attribute Community Search for Semantic Cohesion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for attribute information network community search fail to consider the meaning relations between attributes, leading to insufficient cohesion in community attributes, with potential outcomes of empty or overly large communities.
Innovation Solution
A method and system for attribute community search based on a relation tree, which constructs a relation tree to describe inclusion relations between attributes and defines attribute scores to enhance cohesion, ensuring uniqueness and optimal attribute cohesion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If equal value matching of attributes is used to calculate maximum number of shared attributes, then the calculation is simple, but the community attribute cohesion is insufficient
Solution Approach 1:
The patent changes the parameter of attribute matching from simple equal value comparison to semantic similarity measurement based on Word2Vec vectors. This transforms the matching criterion from binary (match/not match) to continuous (similarity score), enabling more nuanced attribute cohesion while maintaining computational feasibility through vector operations.
Solution Approach 2:
The patent introduces Word2Vec vectors as an intermediary between attributes. Instead of directly comparing attribute values, the system converts attributes to vectors and computes similarity through vector operations. This intermediary layer captures semantic relationships that simple equality matching cannot detect, thereby improving attribute cohesion.
2Speed
If attribute search is performed without considering attribute meaning relations, then the search process is fast, but the community attribute cohesion is low
Solution Approach 1:
The patent performs preliminary action by pre-computing Word2Vec vectors for all attributes in the knowledge base before the actual community search. This preprocessing step stores semantic information in advance, allowing the search phase to quickly retrieve and compare vectors without performing complex semantic analysis during query execution, thus maintaining speed while improving cohesion.
3Adaptability or versatility
If the attributes of the search node are assumed to be related to but not identical to all nodes in the network, then the search is flexible, but an empty community is returned
Solution Approach 1:
The patent changes the matching parameter from strict equality to semantic similarity with a threshold mechanism. By computing similarity scores between attribute vectors and comparing against a threshold, the system flexibly determines attribute relationships. This allows the search to be adaptable to varying degrees of attribute relatedness while ensuring that at least some nodes meet the criteria, preventing empty results.
4Adaptability or versatility
If the attributes of the search node are assumed to be included in all nodes in the entire network, then the search coverage is maximized, but the target community becomes overly large
Solution Approach 1:
The patent applies local quality by using semantic similarity thresholds to create localized filtering at different stages of the search. Instead of uniformly including all potentially related nodes, the system applies similarity-based filtering to identify a focused subset of nodes that truly match the search criteria. This localized quality control prevents the community from becoming overly large while maintaining comprehensive coverage of relevant nodes.
Data Source
AI summary
A method and system for attribute community search based on a relation tree are provided. In the method, a data relation of a target domain is obtained, and an attribute information network of the target domain is constructed in the form of a graph. The attribute information network is loaded over a unified access interface, and data processing, subgraph construction, and algorithm operation and maintenance search are performed on the attribute information network. A community structure parameter k value and a search node q are received through an interface for attribute community search provided for a user, and a community that contains the search node q and satisfies structural cohesion and optimal attribute cohesion is returned. Rendering and interaction are performed on a user interface, which renders the community search result returned into a dynamic community network diagram and presents the dynamic community network diagram to the user.