Cosine similarity Packages

Minimal implementations of a couple of classic text analysis tools (TF-IDF and cosine similarity)

Text classification using n-grams and cosine similarity

Node compatible version of tiny-tfidf

A fork of https://github.com/kerryrodden/tiny-tfidf, removes text splitting and creates 3 ngrams instead.

Inspired by LangMem, nodemem-js is a fast, in-memory vector database for Node.js, designed for efficient similarity search of vector embeddings. Perfect for building chat agent memory and semantic retrieval systems.