cjhutto/vaderSentiment

VADER (Valence Aware Dictionary and sEntiment Reasoner) is a rule-based, lexicon-driven sentiment analyzer that assigns polarity scores to text by matching words against a curated sentiment dictionary and applying linguistic heuristics. It processes text at the sentence level, returning a compound score normalized between -1 (negative) and +1 (positive) along with separate positive, neutral, and negative intensity breakdowns.

What distinguishes VADER from simpler lexicon models is its built-in grammatical rule engine. It adjusts scores for negation (e.g., “not good” reduces positivity), contractions (e.g., “wasn’t” triggers inversion), degree modifiers (e.g., “very” intensifies), and capitalization (uppercase words receive an emotional emphasis boost). The lexicon extends beyond standard dictionary terms to include slang, emoticons (e.g., “:-)“), and acronyms (e.g., “LOL”) common in social media, making it particularly attuned to informal digital communication.

The tool is designed for short-form texts such as tweets, comments, or reviews, but can also aggregate per-sentence scores from longer documents to provide granular emotional tone analysis. It performs text polarity scoring as its primary capability, combining lexical lookup with rule-based adjustments to produce a single normalized compound score. The Python implementation is straightforward to install via standard package managers and can be used directly from the command line or imported as a library for programmatic sentiment analysis.