Scholarlink

This paper proposes a strategic human-in-the-middle framework for linked open data generation in the sociolinguistics domain for the web3.0 environment. It captures human cognition through techniques like social voting, public opinion inclusion, and crowdsourced domain expert knowledge. The framework utilizes TF-IDF for informative term extraction, knowledge store repositories like YAGO and Nell for term enrichment, and E-book metadata generation using BERT. It employs a blend of XG Boost and BERT classifiers for data and metadata classification respectively, balancing computational load. Quantitative semantic reasoning is achieved through APMI measure and cultural search algorithm optimization. The proposed framework achieves high accuracy (97.01%), precision (96.07%), recall (97.95%), and F-measure (97.00%) outperforming baseline models.

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

LODHS: Linked Open Data Generation Using Human-in-the-Middle

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yukta Ramesh,
  • Gerard Deepak

摘要

This paper proposes a strategic human-in-the-middle framework for linked open data generation in the sociolinguistics domain for the web3.0 environment. It captures human cognition through techniques like social voting, public opinion inclusion, and crowdsourced domain expert knowledge. The framework utilizes TF-IDF for informative term extraction, knowledge store repositories like YAGO and Nell for term enrichment, and E-book metadata generation using BERT. It employs a blend of XG Boost and BERT classifiers for data and metadata classification respectively, balancing computational load. Quantitative semantic reasoning is achieved through APMI measure and cultural search algorithm optimization. The proposed framework achieves high accuracy (97.01%), precision (96.07%), recall (97.95%), and F-measure (97.00%) outperforming baseline models.