Question Template Generation in Knowledge Graph QA Systems

Solution Overview

Problem

Current knowledge-graph based artificial intelligence question and answer (QA) systems require significant labor and time to develop query types, question types, and question templates manually, which is inefficient and time-consuming.

Innovation Solution

Automatically generating question templates and question types by parsing graph database queries using a predefined schema, specifically through a query parser that extracts values from the where, return, and match clauses to create natural language question templates.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If question templates are manually developed in KGQA systems, then data accuracy and answer performance are improved, but significant labor and time are required

Engineering Contradiction:

Improvedata accuracyVSAvoidtime to develop query types

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The system performs preliminary action by automatically generating question templates from existing knowledge graph schemas and query patterns before actual QA operations. The template generation module pre-processes the knowledge graph structure to create reusable question templates, eliminating the need for manual template development for each QA interaction while maintaining data accuracy through structured schema-based generation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The KGQA system implements self-service by automatically generating question templates using its own knowledge graph structure and query parsing capabilities. The template generation module extracts patterns from the knowledge graph and autonomously creates question templates without external manual intervention, allowing the system to serve its own template generation needs while preserving accuracy through structured generation.

Inventive Principle:

Principle #25Self-service

2Reliability

If question templates are manually developed in KGQA systems, then answer performance is improved, but the development process is time-consuming

Engineering Contradiction:

Improveanswer performanceVSAvoidtime to develop question types

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary action by automatically generating question templates from existing knowledge graph schemas and query patterns before actual QA operations. The template generation module pre-processes the knowledge graph structure to create reusable question templates, eliminating the need for manual template development for each QA interaction while maintaining data accuracy through structured schema-based generation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The KGQA system implements self-service by automatically generating question templates using its own knowledge graph structure and query parsing capabilities. The template generation module extracts patterns from the knowledge graph and autonomously creates question templates without external manual intervention, allowing the system to serve its own template generation needs while preserving accuracy through structured generation.

Inventive Principle:

Principle #25Self-service

3Productivity

If automated template generation is implemented, then development time is reduced, but manual development ensures high data accuracy

Engineering Contradiction:

Improvetemplate generation speedVSAvoiddata accuracy

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The system applies universality by creating a general-purpose template generation module that can handle multiple query types and knowledge graph schemas through a unified approach. The template generation module uses universal schema patterns and query parsing rules that work across different domains and data structures, enabling high-speed automated generation while maintaining accuracy through consistent structured generation rules.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system applies parameter changes by dynamically adjusting template generation parameters based on the specific knowledge graph schema and query patterns being processed. The template generation module modifies generation parameters such as template structure, variable placeholders, and query mapping rules to match the specific data characteristics, ensuring high accuracy while maintaining automated generation speed through parameter optimization.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11983640B2Generating question templates in a knowledge-graph based question and answer system

Publication Date: 2024.05.14 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Techniques for generating a natural language question template for an artificial intelligence question and answer (QA) system are disclosed. A graph database query relating to a QA system is parsed using a predefined schema. The parsing includes extracting a first plurality of values from the graph database query relating to a where clause in the graph database query, extracting a second plurality of values from the graph database query relating to a return clause in the graph database query, identifying a QA template rule relating to the graph database query, based on a match clause in the graph database query. A natural language question template is generated based on the first plurality of values, the second plurality of values, and the identified QA template rule. The natural language question template is suitable for use by the QA system as part of generating a response to a natural language question.