Scholarlink

<p>Advancements in natural language processing (NLP) have driven research in content evaluation, but no pre-trained models exist for Chinese humanities and social sciences. To address this, several BERT-based pre-trained models for these fields were trained and validated. The results show improved performance in tasks like text classification, structural function recognition, and named entity recognition. These models enhance the intelligent processing of Chinese humanities and social science texts, supporting cross-lingual model development.</p>

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

HsscBERT: pre-training domain model for the full text of Chinese humanity and social science

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月3日

查看全文
  • Si Shen,
  • Zhixiao Zhao,
  • Dayu Yan,
  • Chuan Jiang,
  • Dongbo Wang

摘要

Advancements in natural language processing (NLP) have driven research in content evaluation, but no pre-trained models exist for Chinese humanities and social sciences. To address this, several BERT-based pre-trained models for these fields were trained and validated. The results show improved performance in tasks like text classification, structural function recognition, and named entity recognition. These models enhance the intelligent processing of Chinese humanities and social science texts, supporting cross-lingual model development.