Intelligent question and answer method and system fusing rock knowledge graph

By constructing a rock knowledge graph and employing subject-triggered multi-relation joint extraction and R-GCN model for graph reasoning, the problems of multi-dimensional knowledge fusion and terminology inconsistency in rock question answering are solved, achieving efficient and accurate intelligent question answering, which is applicable to rock classification, engineering attribute query and geological risk assessment.

CN122452791AActive Publication Date: 2026-07-24SHANDONG UNIV +1
11 Cites 2 Cited by

Patent Information

Application Number
CN202610942133.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-29
Publication Date
2026-07-24
Estimated Expiration
2046-06-29

AI Technical Summary

Technical Problem

Existing rock question-answering technologies are limited to the simple inclusion of rock types and basic physical and mechanical properties, lacking multi-dimensional knowledge integration and inconsistent terminology definitions, resulting in weak professionalism, poor practicality, and low accuracy in intelligent question answering.

Method used

Semantic-level processing is performed using a sentence boundary detection model and a pre-trained word vector model to construct a rock knowledge graph. A subject-triggered multi-relation joint extraction mechanism is adopted, combined with the R-GCN model for graph reasoning, and natural language answers are generated through rule templates and text generation models to dynamically update the knowledge graph.

Benefits of technology

It achieves accurate and rapid question answering in the field of rocks, improves retrieval efficiency and accuracy, enhances the professionalism and practicality of the intelligent question answering system, and is suitable for rock classification and identification, engineering attribute query and geological risk assessment.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides an intelligent question-answering method and system fused with a rock knowledge graph, relates to the technical field of crossing of artificial intelligence and geological engineering, and aims to solve the problems that knowledge in the rock field is fragmented, terms are not uniform, the reasoning ability is weak, intelligent question-answering professionality is weak, practicability is poor and accuracy is low. The method comprises the following steps: acquiring multi-source text data, preprocessing the multi-source text data, performing sentence segmentation and standardization processing on the preprocessed multi-source text data, and concurrently extracting a multi-relation triple through a subject trigger type multi-relation joint extraction mechanism; the method comprises the following steps: constructing an ontology structure hierarchy, fusing the ontology structure hierarchy with a multi-relation triple to obtain a rock knowledge graph, analyzing a user question and carrying out graph query, carrying out graph reasoning on graph retrieval content through an R-GCN model by adopting an intention perception mechanism, and generating a natural language answer from a graph reasoning result. The problems of knowledge fragmentation, weak reasoning ability and the like in the rock field are solved, and accurate and rapid question answering in the rock field is achieved.
Need to check novelty before this filing date? Find Prior Art