en-pos
en-pos is a high-accuracy English part-of-speech tagger implemented in JavaScript. Designed for natural language processing tasks, it identifies grammatical tags such as nouns, verbs, adjectives, adverbs, pronouns, prepositions, and punctuation. The library supports installation via NPM and offers a simple API to process token arrays. It utilizes a two-stage approach: an initial dictionary-based tagging phase followed by an optional context-based smoothing step that significantly boosts accuracy. With smoothing enabled, en-pos achieves approximately 96.43 percent accuracy on the Penn Treebank test set, making it one of the most accurate JavaScript taggers available compared to alternatives like pos-js. Disabling the smoothing feature allows for faster processing with slightly lower but still superior accuracy of 94.4 percent. It covers a comprehensive range of tags including singular and plural nouns, proper nouns, various verb tenses and forms, comparative and superlative adjectives and adverbs, determiners,