Finding related posts with embeddings

The useful part

I added a new feature to my blog: a list of related posts at the bottom of each post. I looked at how other content management systems identify related posts: most use shared tags, backlinks, manual curation, or embeddings. I chose embeddings, which compare the meaning of each post, because they can uncover connections without shared tags, existing links, or manual curation.

How it works

• Together, the 768 numbers place the post near other posts with similar meaning.
• During training, the model learns that certain words and phrases appear in similar contexts or play similar roles, so it places them near each other in the space.
• Raw cosine similarity makes everything look related Once every post has an embedding vector, the next question is how to compare them.
• I implemented it using embeddings, and this note documents how.
• Embeddings turn meaning into numbers An embedding model reads text and returns a vector: a long list of numbers.

What to take from it

Conceptually, it is a bit like tagging each blog post with hundreds of auto-generated tags, except that these tags are unnamed (they are just numbers) and distributed (meaning is spread across all of them). It does not need "car" and "automobile" to share any letters to learn that they are used in related ways.

Details worth keeping

This is what lets two posts match even when they use different words. This is where I had to dust off a little math.