This site uses AI for smell information extraction and image generation. Please note that it may contain errors or inaccuracies. Learn more

About This Site

OdGenji: Olfactory Database of The Tale of Genji

This database collects and organizes scent expressions appearing in The Tale of Genji.

Project Overview

This database is part of the results of the National Institute for Japanese Language and Linguistics (NINJAL) collaborative research project "Creating Educational Materials and Building LOD Using the Japanese Classical Text Database and Its Text Data - Through Classroom Practice in Schools and International Collaboration" (Principal Investigator: Masaki Oi, Research Period: 2025-2026, Collaborative Research Number: K454082526).

About the Data

This database contains 140 olfactory descriptions extracted from all 54 volumes of The Tale of Genji.

Source Text:

Data Extraction Method:

  • Automatically extracted from TEI-XML formatted Tale of Genji texts using AI (Google Gemini 2.5 Pro)
  • Strictly extracts only olfactory smell descriptions (excluding visual uses of "nioi" etc.)
  • Includes original text, modern Japanese translation, English translation, plus metadata such as smell quality, location, perceiver, and circumstances

For more details, see the Smell Extraction Guide.

Main Features

  • Smell Search: Search for scent expressions appearing in The Tale of Genji using various criteria
  • Detailed Information: View original text, modern Japanese translation, and English translation for each scent expression
  • Faceted Search: Filter by volume, smell words, quality, location, perceiver, time, and other conditions

Technical Documentation

We provide technical documentation used in building this database:

Open Data

All data is available in RDF/XML format compatible with Semantic Web standards:

Related Links

Contact

For questions or comments about this database, please contact the principal investigator.