Large Language Models for Automatic Knowledge Graph Construction From Text Documents: a Materials Science Case Study
- Vladislav Kaverinskiy(author)
Export Metadata
| Title | Large Language Models for Automatic Knowledge Graph Construction From Text Documents: a Materials Science Case Study |
|---|---|
| DOI | https://doi.org/10.31274/isudp.2025.197.010 |
| Landing page | https://www.iastatedigitalpress.com/plugins/books/197/chapter/1249 |
| Copyright | Vladislav Kaverinskiy |
| Publisher | Iowa State University Digital Press |
| Publication place | Ames, Iowa |
| Published on | 2025-04-14 |
| Long abstract | The presented study investigates the application of large language models (LLMs) in automating the creation of OWL ontologies from scientific articles, focusing on materials science. The research compares the performance of different LLMs, such as GPT-4, Qwen, and DeepSeek, in generating ontologies that capture the essential information and relationships presented in the articles. The study emphasizes the importance of prompt engineering to guide LLMs effectively in this specific domain. Through a series of experiments, the paper demonstrates that while LLMs can produce basic ontologies, there is room for improvement in terms of depth, accuracy, and scalability. The findings suggest that with appropriate prompts and iterative refinement, LLMs have the potential to significantly aid in the creation of knowledge graphs from textual data, thereby enhancing the efficiency and accessibility of scientific information. |
| Language | English (Original) |
