# winkjs/wink-nlp

Developer friendly Natural Language Processing ✨

Repository: https://github.com/winkjs/wink-nlp
Canonical: https://ross.abutalabs.com/products/wink-nlp
Homepage: https://winkjs.org/wink-nlp/
Language: JavaScript
License: MIT
License Family: permissive
Topics: natural-language-processing, nlp, tokenize, sbd, sentence-boundary-detection, negation-handling, sentiment-analysis, pos-tagging, ner, named-entity-extraction, custom-entity-detection, word-vectors, visualization, pattern-matching, wink, wink-nlp, chatbot, hacktoberfest, bm25, vectorizer
Last push: 2026-05-27T02:57:55+00:00

## Health v2 (maintenance only)
Score: 63/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 84, release rhythm 16, longevity 100
- inputs: {"age_days": 2819, "days_push": 98, "days_rel": 429, "gap_med": 108.5, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1388, forks 63 (observed 2026-08-28T04:04:35.233690+00:00)

## What it is
WinkNLP is a developer-friendly JavaScript library for natural language processing with a lean, dependency-free codebase of ~10Kb minified and gzipped. It provides a comprehensive NLP pipeline including tokenization, sentence boundary detection, POS tagging, NER, sentiment analysis, and word embedding support, running on Node.js, browsers, and Deno.

## Use cases
- tokenize multilingual text in javascript
- detect sentences in raw text
- extract named entities from documents
- perform sentiment analysis on reviews
- compute semantic similarity with word vectors
- build a chatbot with nlp in node.js
- create word clouds and text visualizations
- pos tag and lemmatize text in the browser

## When to choose
- you need a fast, lightweight NLP library for JavaScript with no external dependencies
- you want NLP that runs in Node.js, browsers, or Deno with TypeScript support
- you need production-grade text processing with high test coverage and security compliance
- you want tokenization, NER, sentiment, and embeddings in a single pipeline

## When to avoid
- you need deep-learning-based NLP with transformer models or GPU acceleration
- you need languages beyond the models provided (e.g., extensive multilingual pipelines)
- you are working in Python and can use spaCy or similar mature ecosystems

## Facets
- artifact type: library
- maturity: active
- function: nlp, parser, search-engine, data-visualization, machine-learning
- domain: developer-tools, web-development
- platform: browser
- tags: tokenization, sentence-boundary-detection, pos-tagging, named-entity-recognition, sentiment-analysis, word-embeddings, bm25, typescript-support, zero-dependency, natural-language-processing, nodejs, deno, javascript

## Member repositories
- winkjs/wink-nlp (main) score 63

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:35.233690+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:39:47.742829+00:00, confidence not recorded.
  - readme: https://github.com/winkjs/wink-nlp (fetched 2026-08-28T04:04:35.233690+00:00, sha b9500db49074)
  - homepage: https://winkjs.org/wink-nlp/ (fetched 2026-08-29T11:55:28.010726+00:00, sha 350ebf67372d)
  - site_page: https://winkjs.org/wink-nlp/getting-started.html (fetched 2026-08-29T11:55:28.020544+00:00, sha 429ec33b4be4)
  - registry_npm: https://registry.npmjs.org/wink-nlp (fetched 2026-08-29T11:55:28.022395+00:00, sha 48f636c92db3)
- Data as of 2026-08-30T08:39:29.467469+00:00.