Knowledge Graph Implicit Edges via ANN Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Enterprise knowledge graph queries often experience long wait times due to the need to call multiple user-centric services, which can be slow and inefficient, especially when searching for similar documents or information across heterogeneous systems.
Innovation Solution
Incorporating approximate nearest neighbor search as implicit edges in the knowledge graph, allowing for efficient querying by modeling ANN relationships as edges and embeddings as nodes, enabling semantic searches that traverse both explicit and implicit edges in the vector space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If conventional graph query methods are used to search enterprise knowledge graph, then comprehensive search coverage is achieved, but query latency increases significantly
Solution Approach 1:
The system pre-computes and stores embeddings for all entities in the knowledge graph before queries are executed. These embeddings capture semantic relationships and are stored in an ANN index, allowing fast approximate nearest neighbor searches without computing similarities at query time. This preliminary action transforms the expensive query-time computation into a one-time preprocessing step.
Solution Approach 2:
The patent introduces embeddings as an intermediary representation between explicit graph edges and semantic relationships. Instead of directly traversing explicit edges to find semantically similar entities, the system uses embeddings as a mediator that encodes semantic information in vector space, enabling efficient similarity searches through ANN algorithms without requiring exhaustive graph traversals.
2Productivity
If explicit edge traversals are performed to find similar documents, then accurate relationship mapping is achieved, but the number of calls to user-centric services increases
Solution Approach 1:
The patent replaces the mechanical graph traversal process with a mathematical vector space operation. Instead of mechanically following explicit edges through multiple services to find related entities, the system substitutes this with ANN search in embedding space, which uses mathematical distance metrics to efficiently identify semantically similar entities without requiring explicit path traversals.
Solution Approach 2:
The system changes the parameter space from discrete graph edges to continuous vector embeddings. By representing entities and their relationships in a continuous vector space where semantic similarity corresponds to geometric proximity, the system enables efficient similarity searches through distance-based metrics rather than exhaustive edge traversals, fundamentally changing how relationships are queried.
Data Source
AI summary
Systems and methods are directed to incorporating approximate nearest neighbor search as implicit edges in a knowledge graph. The system generates an approximate nearest neighbor (ANN) index that indexes entities by their embeddings. The system models a knowledge graph by including the embeddings as nodes in the knowledge graph. Based on a search query, the system performs a search of the knowledge graph to obtain results, whereby performing the search includes traversing one or more implicit edges from a node of an embedding in the knowledge graph to one or more related nodes in semantic vector space based on the ANN index. The results are then presented on the device of the user.