Triple Verification Device for Knowledge Graph Path Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Knowledge graphs often fail to provide accurate information due to missing relations between entities, as existing embedding methods primarily focus on direct relations and neglect path relations within the graph.
Innovation Solution
A triple verification method that sets a triple with a source entity, target entity, and relation value, extracts intermediate entities, generates connection paths, embeds these paths into vector values, performs convolution operations, applies BiLSTM techniques to generate encoding vectors, and uses an attention mechanism to verify the triple based on similarity values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If embedding methods focus on direct relations between entities, then computational efficiency is improved, but accuracy of information representation deteriorates due to missing path relations
Solution Approach 1:
The patent segments the relation verification process into multiple components: entity embedding, relation embedding, path generation, and verification. By breaking down the complex task of capturing both direct and path relations into separate modules, the system achieves both computational efficiency and accurate information representation. The path generation unit separately identifies intermediate entities and creates multiple paths, allowing efficient processing while capturing comprehensive relation information.
Solution Approach 2:
The patent transitions from two-dimensional direct relation embedding to three-dimensional path-based embedding by introducing intermediate entities and multiple connection paths. This dimensional expansion allows the system to represent not only direct relations but also indirect path relations, improving information representation accuracy while maintaining computational efficiency through structured processing of the additional dimensional data.
2Measurement precision
If multiple connection paths are generated to capture path relations, then accuracy of information representation is improved, but device complexity increases
Solution Approach 1:
The patent creates a universal verification model that handles both direct relations and path relations through the same architectural framework. The verification model takes multiple connection paths as input and processes them uniformly, eliminating the need for separate processing mechanisms for different relation types. This multi-functional approach improves representation accuracy while avoiding the complexity increase that would result from separate specialized modules.
Solution Approach 2:
The patent merges the processing of direct relations and path relations into a unified verification framework. By combining multiple connection paths (including direct and indirect relations) into a single input structure for the verification model, the system achieves accurate representation of both relation types without requiring separate processing pipelines, thus avoiding excessive system complexity.
3Reliability
If path relations are inferred through intermediate entities, then quality of knowledge graph improves, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by pre-extracting intermediate entities and pre-generating connection paths before the main verification process. The path generation unit identifies intermediate entities and creates multiple paths in advance, so that during verification, the system only needs to process these pre-computed paths rather than discovering relations in real-time. This preliminary processing improves knowledge graph quality while reducing actual verification processing time.
Solution Approach 2:
The patent maintains continuity of useful action by using the same embedding vectors and path representations across multiple verification operations. The entity and relation embeddings are computed once and reused, while path representations are generated consistently, allowing efficient batch processing. This continuous reuse of computational results improves knowledge graph quality through comprehensive path analysis while minimizing redundant processing time.
Data Source
AI summary
A triple verification method is provided. The triple verification method includes setting a triple having a source entity, a target entity, and a relation value between the source entity and the target entity by a setting unit, extracting a plurality of intermediate entities associated with the source entity and the target entity by the setting unit, defining a connection relation between the intermediate entity, the source entity, and the target entity and generating a plurality of connection paths connecting the source entity, the intermediate entity, and the target entity by a path generation unit, generating a matrix by embedding the plurality of connection paths into vector values by a first processing unit, calculating a feature map by performing a convolution operation on the matrix by a second processing unit, generating an encoding vector for each connection path by encoding the feature map by applying a bidirectional long short-term memory neural network (BiLSTM) technique by a third processing unit, and generating a state vector by summing the encoding vectors for each connection path by applying an attention mechanism and verifying the triple based on a similarity value between the relation value of the triple and the state vector by a determination unit.