Knowledge Graph Indexing with Informative Triplet Filtering

Solution Overview

Problem

Incorporating non-informative data into knowledge graphs increases storage requirements and reduces the efficiency and effectiveness of data retrieval, leading to irrelevant search results.

Innovation Solution

Generate document chunks and summarizations from documents, determine entity types and relations, create a schema, and build a knowledge graph with informative entity property and relation triplets, allowing for efficient retrieval by traversing the graph based on search queries.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If non-informative data is incorporated into the knowledge graph, then the knowledge graph contains more data, but storage space increases and retrieval efficiency decreases

Engineering Contradiction:

Improvedata volumeVSAvoidretrieval efficiency

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The patent extracts and removes non-informative data from the knowledge graph construction process. By identifying and excluding data that does not contribute to meaningful entity relationships, the system maintains a compact knowledge graph that optimizes retrieval efficiency while preserving only the essential informative data.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies different quality standards to different portions of data during knowledge graph construction. Informative data that contributes to entity relationships is retained with high quality, while non-informative data is filtered out. This selective quality approach ensures optimal storage efficiency and retrieval performance.

Inventive Principle:

Principle #3Local quality

2Quantity of substance

If non-informative data is incorporated into the knowledge graph, then the knowledge graph contains more data, but storage space requirements increase

Engineering Contradiction:

Improvedata volumeVSAvoidstorage space

Core Design Contradiction:

Quantity of substanceVSWeight of stationary object

Solution Approach 1:

The patent extracts and removes non-informative data from the knowledge graph construction process. By identifying and excluding data that does not contribute to meaningful entity relationships, the system maintains a compact knowledge graph that optimizes retrieval efficiency while preserving only the essential informative data.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies different quality standards to different portions of data during knowledge graph construction. Informative data that contributes to entity relationships is retained with high quality, while non-informative data is filtered out. This selective quality approach ensures optimal storage efficiency and retrieval performance.

Inventive Principle:

Principle #3Local quality

3Quantity of substance

If non-informative data is incorporated into the knowledge graph, then more documents are returned for search queries, but the relevance of returned documents decreases

Engineering Contradiction:

Improvenumber of returned documentsVSAvoiddocument relevance

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent extracts and removes non-informative data from the knowledge graph construction process. By identifying and excluding data that does not contribute to meaningful entity relationships, the system maintains a compact knowledge graph that optimizes retrieval efficiency while preserving only the essential informative data.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies different quality standards to different portions of data during knowledge graph construction. Informative data that contributes to entity relationships is retained with high quality, while non-informative data is filtered out. This selective quality approach ensures optimal storage efficiency and retrieval performance.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20260030517A1Efficient knowledge graph indexing and retrieval

Publication Date: 2026.01.29 SALESFORCE INC

AI summary

Systems, devices, and techniques are disclosed for efficient knowledge graph indexing and retrieval. Document chunks may be generated from documents. Summarizations may be generated from document chunks. Entity types, entity properties, relations, and relation properties may be generated from a subset of the summarizations. A schema including entity types, entity properties, relations, and relation properties may be generated. Entity property triplets and entity relation triplets may be generated from the summarizations based on the schema and linked to the document chunks. A knowledge graph including nodes representing entities from the entity property triplets and entity relation triplets and edges representing the entity property triplets and the entity relation triplets may be generated. A search query may be received. Nodes and edges of the knowledge graph that include the entities, the entity property triplets and the entity relation triplets most similar to keywords of the search query may be determined.