Implement the `--lsi` option to show related posts

[osteele]

Jekyll uses https://github.com/jekyll/classifier-reborn

For Go, look at https://github.com/avelino/awesome-go/#machine-learning, in particular https://github.com/jbrukh/bayesian.

GitHub search also turned up https://github.com/AlasdairF/Classifier.

[osteele]

Implementation Analysis for --lsi Option

The infrastructure is already partially in place:

• related_posts variable is exposed to Liquid templates
• Tests exist for related_posts
• Currently returns the 10 most recent posts (no semantic analysis)

Algorithm: TF-IDF + Cosine Similarity

Advantages:

• No external dependencies (pure Go implementation)
• Fast for typical blog sizes
• Industry-standard for document similarity

Design:

1. Extract text from each post (title + content + tags)
2. Tokenize and remove stop words
3. Build TF-IDF vectors for each document
4. Calculate cosine similarity between posts
5. Return top 10 most similar (excluding self)