Scholarlink

The traditional format of academic articles, while optimized for human readability, poses challenges for automated systems to process and integrate scientific knowledge. These problems are more serious for low-resource languages like Kazakh, which lack structured and comprehensive information resources. In this study, we built multilingual knowledge graphs (KGs) of natural language processing field for Kazakh, English, and Russian to solve this problem. Using these KGs, we developed a multilingual information resource system that facilitates access to structured knowledge across languages. The resulting system enhances interoperability and bridges the resource gap for Kazakh, while supporting multilingual scientific knowledge integration and retrieval.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automatic Creation of Multilingual Knowledge Graph with Large Language Models

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Gulmira Tolegen,
  • Alymzhan Toleu,
  • Rustam Mussabayev,
  • Alexander Krassovitskiy,
  • Nurbakyt Zhuldyzbayuly

摘要

The traditional format of academic articles, while optimized for human readability, poses challenges for automated systems to process and integrate scientific knowledge. These problems are more serious for low-resource languages like Kazakh, which lack structured and comprehensive information resources. In this study, we built multilingual knowledge graphs (KGs) of natural language processing field for Kazakh, English, and Russian to solve this problem. Using these KGs, we developed a multilingual information resource system that facilitates access to structured knowledge across languages. The resulting system enhances interoperability and bridges the resource gap for Kazakh, while supporting multilingual scientific knowledge integration and retrieval.