An 18k-star NLP index that leaves chatbots to other lists

What it does

This is a curated index of roughly 18,000-starred resources spanning natural language processing. It organizes libraries across a dozen languages, research archives, tutorials, annotation tools, task-specific methods, multilingual frameworks, and per-language corpora into a single hierarchical reading list. The maintainer enforces a narrow scope: large language models appear only when they advance a core NLP capability like tokenization, translation, or factuality, while general-purpose chatbots and prompt-template repositories are explicitly sent elsewhere.

The interesting bit

The list’s value is as much about what it excludes as what it includes. In an era where every repository is rebranding as an “AI agent framework,” this index still makes room for endangered-language MT systems, phonetic representation modeling, and the Penn Treebank. That curatorial stubbornness is rare.

Key highlights

• Covers libraries in Python, C++, Java, Rust, Go, R, Scala, Kotlin, Clojure, Ruby, Julia, Node.js, and NLP++.
• Sections for 17+ individual languages including Arabic, Chinese, Indic languages, Korean, Persian, and Ukrainian.
• Explicitly scopes out general chatbots, agent frameworks, code-generation tools, and RAG application starter kits.
• Includes historical deep cuts like the Avenue Project for endangered languages and the Penn Discourse Treebank.
• Maintains dedicated sections for classical tasks: POS tagging, coreference resolution, topic modeling, and morphological segmentation.

Verdict

NLP researchers, linguists, and engineers tired of wading through LLM chatbot repositories to find a solid Arabic NER dataset or a morphological segmentation library should bookmark this. If you are looking for a RAG starter kit or the latest prompt-engineering cheat sheet, the README itself tells you to look elsewhere.