Sentibank: A Unified Resource of Sentiment Lexicons and Dictionaries

Abstract

Sentiment analysis is critical across computational social science domains, but faces challenges in interpretability. Rule-based methods relying on expert lexicons enable transparency, yet applying them is hindered by resource fragmentation and lack of validation. This paper introduces sentibank, a large-scale unified database consolidating 15 original sentiment dictionaries and 43 preprocessed dictionaries, spanning 7 genres and 6 domains.