- Hugo websites use tags to organize blog posts into taxonomies, but tags may be underutilized by users.
- Affiliation networks are better suited for Hugo sites than Obsidian-style graphs because posts are linked via tags, not cross-links.
- An affiliation network is an undirected bipartite graph with two disjoint node sets: blog posts and tags.
- The author built the graph using Python with BeautifulSoup and NetworkX, extracting tags from HTML files and outputting JSON.
- Tags are processed as part of a GitHub Action to update the graph automatically on site builds.
- D3.js is used for visualization, with force-directed layout, color-coded nodes (blue for posts, pink for tags), and labels with links.
- Nodes are clickable: blog posts link to their pages, tags link to tag listing pages.