Rayford GeoGraph

Open Live Website | Make Your Own | Google Scholar | Main Homepage | 中文说明

Rayford GeoGraph is my personal contribution to a more open research knowledge layer: a public knowledge base and knowledge graph for my papers, book chapters, datasets, GitHub repositories, and GeoAI workflows.

Instead of presenting research as a flat publication list, this project treats each output as a connected object with a paper trail, code trail, method trail, and intellectual lineage. The goal is to make my research easier to inspect, reuse, teach, and extend.

What This Is

• A public research atlas for Yifan Yang's GeoAI and GIScience work.
• A star-map style website for exploring papers, repositories, datasets, and methods.
• A structured markdown knowledge base that agents and humans can maintain together.
• A weekly-updated Google Scholar snapshot for public citation metadata.
• A forkable template for other researchers who want to build their own research graph.

One-Click Access

• Live site: rayford295.github.io/Rayford-GeoGraph
• Fork guide: rayford295.github.io/Rayford-GeoGraph/fork.html
• GitHub repository: github.com/rayford295/Rayford-GeoGraph
• Google Scholar: scholar.google.com/citations?user=B-fiSHwAAAAJ
• Main academic homepage: rayford295.github.io

Frontend Experience

• The first screen is an interactive research constellation.
• The graph supports keyword search, theme filters, repository cards, and three views: Network, Timeline, and Repo.
• Each node opens an inspector with repository metadata, methods, paper links, dataset links, and graph relationships.
• The visual language is intentionally celestial: research outputs appear as connected stars in the same research sky.

Knowledge Architecture

• wiki/papers/: structured profiles for research outputs.
• wiki/concepts/: reusable concept pages.
• wiki/comparisons/: cross-paper research narratives.
• raw/papers/: source records for paper and repository metadata.
• raw/scholar/google-scholar.json: the latest Google Scholar profile snapshot.
• scripts/build-map.js: compiles paper pages into data.js.
• scripts/fetch-scholar.js: refreshes public Google Scholar metadata.
• .github/workflows/update-scholar.yml: runs the Scholar refresh once per week.
• docs/FORK_GUIDE.md: explains how another researcher can fork and customize the atlas.
• wiki/papers/_template.md: reusable paper-node template ignored by the graph build.

Current Research Outputs

• ArcGIS Text SAM Tree Segmentation
• GeoLocator
• Hyperlocal Disaster Damage Assessment
• Perceiving Multidimensional Disaster Damages
• DamageArbiter
• Satellite-to-Street

Weekly Scholar Refresh

The repository includes a scheduled GitHub Actions workflow that runs once per week. It fetches the public Google Scholar profile, updates raw/scholar/google-scholar.json, and commits the new snapshot only when the data changes.

Google Scholar may temporarily rate-limit automated requests. When that happens, the script keeps the previous snapshot and avoids breaking the website.

Local Workflow

npm run build
npm run scholar:update

Use npm run build after editing wiki/papers/. Use npm run scholar:update when you want to refresh the local Google Scholar snapshot manually.

Make Your Own

This repository is designed to be forked. Start with the Make Your Own page, then follow docs/FORK_GUIDE.md. The short version:

1. Fork and rename the repository.
2. Replace personal links, Scholar user id, and homepage metadata.
3. Copy wiki/papers/_template.md for each research output.
4. Run npm run build.
5. Enable GitHub Pages and publish your own research graph.

Next Growth Directions

1. Add dedicated project pages for each research output.
2. Expand wiki/concepts/ and wiki/comparisons/.
3. Add talks, datasets, code releases, and collaborators as graph entities.
4. Add bilingual content blocks directly on the website.