Chemical Knowledge Graph Construction via NLP Entity Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The chemical industry faces challenges in managing the wide range and complexity of knowledge, leading to difficulties in handling emergencies and making decisions due to the lack of effective intelligent question answering technologies specifically tailored for the field, with existing solutions failing to understand user needs and provide accurate solutions.
Innovation Solution
A method and device for constructing a chemical knowledge graph using natural language processing, big data, and artificial intelligence, which involves preprocessing and aligning data to create a standard knowledge representation, enabling efficient knowledge inference and question answering within the chemical field.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing retrieval technology or deep learning matching technology is used for intelligent question answering, then general question answering capability is provided, but the system cannot efficiently respond to the wide range, variety and large quantity of knowledge in the chemical industry and cannot understand user needs
Solution Approach 1:
The patent introduces a chemical knowledge graph as an intermediary structure between the question answering system and chemical industry knowledge. The knowledge graph contains standardized chemical entities, relationships, and properties that serve as a mediator to bridge general NLP capabilities and domain-specific chemical knowledge, enabling the system to understand and respond to chemical industry questions effectively
Solution Approach 2:
The patent performs preliminary action by automatically constructing and pre-processing the chemical knowledge graph before deployment. This includes entity recognition, relationship extraction, and knowledge graph construction from chemical industry data sources, so that when users ask questions, the system already has structured chemical knowledge ready for efficient querying and understanding
2Manufacturing precision
If manual construction method is used for chemical knowledge graph, then high accuracy and completeness can be achieved, but the construction speed is slow and manual construction cost is high
Solution Approach 1:
The patent replaces the mechanical manual construction process with automated computational methods. It uses natural language processing, entity recognition algorithms, and knowledge graph construction algorithms to automatically extract chemical entities, relationships, and properties from data sources, substituting human manual work with automated systems that achieve both speed and accuracy
Solution Approach 2:
The system performs self-service by automatically constructing and maintaining the chemical knowledge graph without requiring manual intervention for each update. The automated pipelines continuously crawl, extract, and integrate new chemical knowledge from data sources, allowing the knowledge graph to self-upgrade and maintain accuracy while improving construction speed
Data Source
AI summary
The disclosure relates to a construction method and device for a chemical knowledge graph, an intelligent question answering method and device for chemical knowledge, and two computer-readable storage media. The construction method comprises the following steps: obtaining knowledge data in chemical industry field; pre-processing the knowledge data to obtain entity data and property data related to chemical knowledge; determining a preliminary knowledge representation according to the entity data and the property data; performing entity alignment on the preliminary knowledge representation to obtain a standard knowledge representation; and constructing the chemical knowledge graph according to the standard knowledge representation. The construction method for a chemical knowledge graph can automatically collect relevant knowledge in the chemical industry to construct a chemical knowledge graph on basis of natural language processing, big data and artificial intelligence technology, thereby greatly improving the construction speed of the chemical knowledge graph and reducing the manual construction cost of the chemical knowledge graph.