Knowledge Graph Augmentation via ML Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Knowledge graphs often contain incomplete or missing relationship information, which can lead to inaccurate or incomplete responses when querying the graph, as users may provide partial specifications of new relationship triples without sufficient information to fill in the target entity or relationship type.

Innovation Solution

A computer-implemented method that builds a dataset from an initial knowledge graph, using machine learning models such as clustering and classification models to predict missing relationship information, allowing for the augmentation of the knowledge graph with complete relationship triples by identifying the missing information and suggesting it to the user.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual completion of relationship triples is performed, then accuracy of relationship information can be ensured, but time consumption and operational complexity increase significantly

Engineering Contradiction:

Improveaccuracy of relationship informationVSAvoidtime consumption for completing relationship triples

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables automatic completion of relationship triples by having the knowledge graph system itself perform the completion task using machine learning models, eliminating the need for manual intervention while maintaining accuracy through intelligent prediction of missing target entities and relationship types

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical operations with automated machine learning-based systems. Clustering models and classification models substitute human analysts, automatically predicting missing relationship information based on patterns learned from existing knowledge graph data

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If machine learning models are used to predict missing relationship information, then productivity and automation are improved, but system complexity increases

Engineering Contradiction:

Improvespeed of knowledge graph augmentationVSAvoidcomplexity of machine learning system

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system divides the complex task of relationship triple completion into two separate machine learning models: a clustering model that predicts target entities and a classification model that predicts relationship types. This segmentation reduces the complexity of each individual model while maintaining overall system effectiveness

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediate processing steps where the clustering model first identifies potential target entities, and then the classification model determines the relationship type. This intermediary approach simplifies the overall prediction process by breaking it into manageable stages with clear inputs and outputs

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If complete relationship triples are required for all queries, then query accuracy is improved, but the effort and time to obtain complete information increases

Engineering Contradiction:

Improveaccuracy of query responsesVSAvoidease of adding relationship information

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system performs preliminary action by automatically completing relationship triples in advance before queries are executed. The machine learning models predict and fill in missing information proactively, so that when queries are run, the knowledge graph already contains complete relationship triples, ensuring accuracy without requiring manual completion at query time

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20240403664A1Determining missing relationship information and augmenting a knowledge graph

Publication Date: 2024.12.05 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Determining missing relationship information and augmenting a knowledge graph includes obtaining a knowledge graph with nodes representing entities and edges representing relationships between related entities, building a dataset indicating attribute(s) and attribute value(s) of each entity of a related entity pair, a relationship type of the relationship between the related entity pair, and a weight of the relationship type, using the dataset to build machine learning model(s), receiving a partial specification of a new relationship triple, the specification including the new source entity and missing relationship information of the new relationship triple, applying one or more of the model(s) and identifying the missing relationship information, and augmenting the graph to provide an augmented knowledge graph that includes the new relationship triple, including the new source entity, the target entity, and the relationship type of the relationship between the new source entity and the target entity.