find47

日本語のREADMEはこちらです: README.ja.md

This repository provides an open dataset of scenic photos from across Japan, sourced from the FIND/47 project. It includes scripts to automatically scrape and update the data, as well as sample applications to demonstrate its use.

Demos

• Photo List: An endless-scrolling gallery of all photos.
• Photo Search: Search for photos by keyword.
• Map: View all photos on an interactive map.
• Nearby Scenic Spots Search: Find scenic spots near your current location.

Features

• Comprehensive Datasets: Available in CSV and JSON, with full and compact versions. Includes metadata like titles, authors, EXIF data, and geolocation.
• High-Quality Images: A collection of photos from all 47 prefectures of Japan, stored in the photo/ directory.
• Automated Updates: Scripts to periodically check for new images and update the dataset.
• Ready-to-Use Demos: Includes web apps for searching, mapping, and discovering photos.
• Open License: All images are licensed under CC BY 4.0, making them easy to reuse with proper attribution.

Datasets

The data is available in both full and compact (lightweight) versions.

• Full version:
  • find47images.csv
  • find47images.json
• Compact version:
  • find47images_c.csv
  • find47images_c.json

Usage Example

You can fetch the compact JSON data directly in your web application:

const data = await (await fetch("https://code4fukui.github.io/find47/find47images_c.json")).json();
const photo = data.find((d) => d.id == 4034);
console.log(photo);

Data Schema

The compact dataset (find47images_c.csv) contains essential fields for lightweight applications:

• id: Unique identifier for the image.
• code: Shortcode used by FIND/47.
• pref: Prefecture in Japanese.
• title: Title in Japanese.
• author: Author's name in Japanese.
• Geo3x3: Geolocation code.

The full dataset (find47images.csv) includes additional fields such as English translations (title_en, author_en), URLs, view/download counts, and detailed EXIF data (camera, lens, date shot, etc.).

License and Attribution

The program code in this repository is licensed under the MIT License.

The images in photo/ and their metadata are distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license by the authors listed in the author / author_en fields.

When using or redistributing an image, you must provide proper attribution. Please include at least:

• Author/Provider: author or author_en
• License: https://creativecommons.org/licenses/by/4.0/
• Source Page: url or url_en (when reasonably practicable)
• Indication of modifications, if any.

Including the title (title or title_en) is also recommended for easier identification.

Example Attribution:

"Cape Notoro" by Masatoshi Konishi, FIND/47, CC BY 4.0
https://find47.jp/en/i/NzJVV
https://creativecommons.org/licenses/by/4.0/

For Developers

Data Update Process

The dataset is kept up-to-date using a series of Deno scripts:

1. download.js: Scrapes find47.jp to get the latest image metadata and generates find47images.csv.
2. downloadPhoto.js: Downloads the corresponding image files into the photo/ directory.
3. makeCompact.js & makeJSON.js: Generate the compact and JSON versions of the dataset.

To run the update process:

deno run -A download.js
deno run -A downloadPhoto.js

Utilities

• deleteByAuthor.js: A command-line tool to remove all data and images associated with a specific author. This is useful for processing data removal requests.

Sample Applications

This dataset is used in other projects, such as:

• Saigaikun
• Fukui Tourism App