Knowledge Graph Reasoning for Complex Query Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated question and answer systems are unable to handle questions requiring logical inference and understanding of context, limiting their ability to provide accurate responses to complex queries in domains like accounting and finance.
Innovation Solution
The development of systems and methods that construct knowledge graph empowered question-and-answering knowledge bases, capable of performing inferential logical reasoning. These systems extract topic entities, group them into clusters, identify linguistic modalities, and construct data structures like knowledge graphs to process and generate responses to complex queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing automated question and answer systems are used, then definition-type questions can be answered, but logical inference type questions cannot be answered
Solution Approach 1:
The system segments the question answering process into distinct modules: intent detection module that identifies question type (definition vs. logical inference), knowledge graph construction module that builds structured representations, and reasoning module that applies inference rules. This segmentation allows the system to handle different question types through specialized pathways, enabling both definition-type and logical inference type questions to be answered reliably
Solution Approach 2:
The patent introduces a knowledge graph as an intermediary structure between the input question and the answer generation process. The knowledge graph captures entities, relationships, and logical rules from domain knowledge, serving as a mediator that enables logical inference. When a logical inference question is detected, the system queries the knowledge graph to derive answers through reasoned paths rather than direct text matching
2Adaptability or versatility
If knowledge graph empowered systems are constructed, then logical inference capability is improved, but system complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-construction of knowledge graphs from domain knowledge sources before actual question answering occurs. The knowledge graph is built in advance with entities, relationships, and logical rules encoded, so that during query processing, the system only needs to query and reason over the pre-structured knowledge rather than constructing it from scratch for each question. This reduces real-time computational complexity while maintaining inferential capability
Solution Approach 2:
The system dynamically adapts its processing pathway based on the detected question type. For definition-type questions, it uses direct text retrieval; for logical inference questions, it activates the knowledge graph reasoning pathway. This dynamic switching optimizes resource usage and manages complexity by only engaging the full knowledge graph infrastructure when necessary for inferential tasks
Data Source
AI summary
An exemplary system for constructing data structures that can perform inferential reasoning to answer input queries may receive input data, extract and cluster entities in the input data into topic clusters, and for a first topic cluster construct a data structure comprising a plurality of nodes, wherein nodes of the data structure respectively represent a topic entity extracted from the input data and grouped into the first topic cluster, and wherein a first node of the data structure is associated with a second node of the data structure based on the first node and the second node respectively representing a first topic entity and a second topic entity associated in the input data with a common one of the one or more identified linguistic modalities. An exemplary system comprising the data structure may receive an input query and generate a response to the input query using the data structure.