Scholarlink

<p>This paper reports MuSET VGG, a hybrid ensemble architecture system for translation from classical Sanskrit meteorological texts and extraction of validated rainfall events. The system integrates rule-based, statistical, neural, and symbolic translation components into a single architecture guided by a domain-specific validation layer. Unlike existing models, MuSET VGG implements gated attention fusion which is coupled with ontology-guided temporal logic process. Both of these are used to guarantee semantic fidelity and temporal consistency in translation outputs. The foundation is a parallel corpus of Meghamala verses and a structured cosmological ontology. The experimental results reveal marked improvements in BLEU, METEOR, and domain-specific metrics for different scenarios. These include event extraction accuracy and time alignment sets. The framework here constitutes a major advancement in the translation of Sanskrit–English by providing interpretive access to environmental knowledge embedded in ancient texts.</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

MuSETVGG: A Hybrid Ensemble Framework for Sanskrit–English Translation and Rainfall Prediction from Meghamala Texts

  • 期刊论文

  • 发表日期：

    2025年10月25日

查看全文
  • Pradnya Patil,
  • Sunita Patil

摘要

This paper reports MuSET VGG, a hybrid ensemble architecture system for translation from classical Sanskrit meteorological texts and extraction of validated rainfall events. The system integrates rule-based, statistical, neural, and symbolic translation components into a single architecture guided by a domain-specific validation layer. Unlike existing models, MuSET VGG implements gated attention fusion which is coupled with ontology-guided temporal logic process. Both of these are used to guarantee semantic fidelity and temporal consistency in translation outputs. The foundation is a parallel corpus of Meghamala verses and a structured cosmological ontology. The experimental results reveal marked improvements in BLEU, METEOR, and domain-specific metrics for different scenarios. These include event extraction accuracy and time alignment sets. The framework here constitutes a major advancement in the translation of Sanskrit–English by providing interpretive access to environmental knowledge embedded in ancient texts.