Home/Datasets/Temples and Ancient Architecture Dataset

Temples and Ancient Architecture Dataset

The Temples and Ancient Architecture Dataset is a curated collection of Odia text by the Odia Virtual Academy (OVA). It covers temple architecture, sacred structures, sculptural art, and heritage conservation, including Kalinga style, stone carvings, inscriptions, and restoration practices. Designed for language modelling and NLP research, the dataset supports domain adaptation, information extraction, and AI-assisted education in Odia for architectural heritage.

About Dataset

The Temples and Ancient Architecture Dataset is a meticulously curated collection of Odia text by the Odia Virtual Academy (OVA). It brings together material across temple studies, architectural history, archaeology, and heritage conservation to reflect the breadth and complexity of these interrelated domains. The dataset encompasses topics such as temple design principles and structural systems, architectural styles and typologies, stone carving and sculptural traditions, iconography and symbolism, epigraphy and inscriptions, ritual spaces, and construction techniques, as well as regional variations including Kalinga architecture and other classical traditions of India. It further covers heritage management, conservation practices, documentation of monuments, cultural landscapes, pilgrimage traditions, historical narratives, and traditional knowledge embedded in Odia-speaking communities. Curated by Odia Virtual Academy, the dataset emphasizes domain-relevant terminology, regional expressions, and architecture-specific vocabularies to support accurate language modelling, information extraction, and domain adaptation for Odia NLP research and language modelling.

Purpose of Dataset

The Temples And Ancient Architecture Dataset Aims To Provide Odia Text Sourced From Digitized Temple Records, Architectural Treatises, Epigraphic Sources, And Heritage-related Materials. It Is Suitable For A Range Of Applications, Including Training Language Models, Building Domain-aware Nlp Tools (Such As Named-entity Recognition For Temples, Monuments, Architectural Elements, Deities, Inscriptions, And Historical Periods; Relation Extraction; And Multilingual Grounding), Implementing Ai-assisted Education And Heritage Awareness For Students And Researchers, And Enabling Content Generation That Aligns With Temple Studies And Architectural History Contexts In Odia.