Entity Combination Graph Nodes for Sparse Data Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large, sparse matrices and graph databases in distributed computer environments face challenges with scalability and inefficient storage, making it difficult to monitor and analyze data, as they require significant storage space and are cumbersome to traverse or search.
Innovation Solution
A directed acyclic graph is used to represent entity combinations as nodes, allowing for efficient storage and parsing by linking nodes with nearest neighbor nodes, reducing the need for multiple columns and storing only used combinations, and enabling quick identification of relevant entities and combinations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If large, sparse matrices and graph databases are used to store entity combinations in distributed computer environments, then complete data coverage is achieved, but storage requirements increase significantly and data traversal becomes inefficient
Solution Approach 1:
The patent extracts and stores only the essential information needed for efficient querying - specifically, the nearest neighbor relationships between entity combinations. Instead of storing complete sparse matrices with all possible entity combinations, the system extracts only the meaningful connections (nearest neighbors) and stores them in a compressed graph structure, thereby reducing storage requirements while maintaining data coverage.
Solution Approach 2:
The patent segments the large-scale entity combination data into smaller, manageable graph components. Each entity combination is represented as a node, and the data is organized into a graph structure where nodes are connected through nearest neighbor relationships. This segmentation allows the system to handle large datasets efficiently by processing and traversing localized graph regions rather than managing monolithic sparse matrices.
2Reliability
If traditional graph databases are used to store entity combinations, then data relationships are preserved, but traversal and searching become cumbersome and inefficient
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing the nearest neighbor relationships between entity combinations during the graph construction phase. Instead of computing relationships on-demand during traversal (which would be inefficient), the system performs the computationally intensive nearest neighbor calculations in advance and stores the results in the graph structure. This preliminary preparation enables fast traversal and searching operations during query execution.
3Quantity of substance
If multiple columns are used to represent every possible service combination, then complete combination coverage is achieved, but storage efficiency decreases and parsing becomes more complex
Solution Approach 1:
The patent applies universality by using a single graph data structure that can represent any entity combination through its nodes and relationships. Instead of creating separate columns or data structures for each possible service combination, the graph structure universally handles all combinations by representing them as nodes with appropriate labels and connections. This multi-functional approach allows the same structure to efficiently represent sparse or dense combination data without requiring parsing complexity to increase.
Data Source
AI summary
A graph of combinations of entities and parameters corresponding to the combinations of entities may be stored as two tables. The first table may comprise a table that includes all entity combinations, as well as each parameter that corresponds to the entity combinations. Each entity combination may additionally be parseable, such that each entity combination may be parsed to allow for identification of each entity included within a given entity combination. The second table may include an entity combination node corresponding to (and linked to) each entity combination stored within the first table. Each given entity combination node of the second table may then be linked within the second table to each nearest neighbor node of the given node to thereby allow for identifying each entity combination within the first table that includes a particular relevant entity (or set of entities).