Knowledge Graph Search via Vector Traversal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for searching digital objects in knowledge graphs are inefficient, particularly in large document collections, as they struggle to accurately retrieve relevant content units and highlight key entities within the search results.
Innovation Solution
A computer-implemented method for searching a graph representing digital objects, which involves receiving a search request, determining a set of operations for traversing the graph, executing these operations to generate intermediate and result vectors of nodes, retrieving relevant object units, selecting intermediate result vectors, identifying result entities, and providing these entities and object units as search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional search methods are used on large document collections, then the search process can be simple to implement, but the retrieval accuracy and efficiency deteriorate
Solution Approach 1:
The patent segments the search process into distinct operational steps (determining operations, executing operations to generate vectors, selecting intermediate vectors, identifying entities) and segments the data structure into nodes representing entities and edges representing relationships. This segmentation allows the system to process large document collections efficiently by breaking down the complex search task into manageable operations that can be executed systematically.
Solution Approach 2:
The patent transforms the search problem from traditional text-matching dimensions into a graph-based dimensional space where digital objects are represented as nodes and relationships as edges. By converting search results into vectors of nodes and performing operations in this graph dimension, the system achieves both high retrieval accuracy through relationship analysis and improved efficiency through structured graph traversal operations.
2Measurement precision
If comprehensive graph traversal operations are performed to ensure accurate entity identification, then the retrieval accuracy improves, but the computational complexity and time consumption increase
Solution Approach 1:
The patent applies preliminary action by pre-defining a structured set of operations for graph traversal before executing the search. These operations are determined in advance based on the search request and the graph structure, allowing the system to systematically process the knowledge graph without ad-hoc complexity during execution. This preliminary structuring reduces computational complexity while maintaining accurate entity identification.
Solution Approach 2:
The patent introduces intermediate result vectors as mediators between the graph traversal operations and the final entity identification. These intermediate vectors serve as a structured intermediary representation that simplifies the complexity of direct graph analysis while preserving the accuracy needed for precise entity identification. The intermediary vectors organize the traversal results in a manageable format for subsequent processing.
3Reliability
If the system retrieves and processes all potential result vectors to ensure completeness, then the comprehensiveness of results improves, but the retrieval time and resource consumption worsen
Solution Approach 1:
The patent applies partial action by selectively processing intermediate result vectors rather than all possible vectors. The system determines a specific set of operations to execute and selectively processes intermediate vectors that are most relevant to the search request. This approach maintains result completeness for the most important entities while reducing retrieval time by avoiding unnecessary processing of all potential vectors.
Data Source
AI summary
The present disclosure relates to a method for searching a graph representing content of digital objects. A set of operations for traversing the graph may be determined according to a search request. The set of operations may be executed, resulting in intermediate result vectors of nodes and a result vector of nodes, wherein the result vector of nodes is associated with a result set of one or more object units of the digital objects. Intermediate result vectors may be selected from of the intermediate result vectors. A set of result entities may be identified. The set of result entities are entities which are part of the object units and part of entities represented by nodes of said selected intermediate result vectors. The set of result entities and the result set of object units may be provided as a result of the search request.