EUVIMEDEUVIMED entdecken
Crossref

EmojiSentR: An R Package for Integrated Text and Emoji Sentiment Analysis

Logan Hanson, Elias Lahrim, Xin Tong

Journal of Behavioral Data Science · 30.03.2026 · 6 (1) · 152-164

DOI: 10.35566/jbds/tonghl ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Behavioral Data Science, online erschienen am 30.03.2026. Verfasst von Logan Hanson und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Although emojis are used by 92% of the world’s online population, appear in over one quarter of social media posts, and carry affect beyond words, most software packages for sentiment analysis ignore emoji semantics. Our newly developed R package, EmojiSentR, addresses this gap by merging emoji valence with word-level sentiment in a tidy R workflow. The package bundles an internal lexicon derived from the Novak-1200 dataset (1,200 glyphs 8 emotions) and a pipeline function, sentiment_analysis(), that (1) extracts emojis, (2) cleans residual text, (3) scores each channel separately, (4) leverages VADER’s (Valence Aware Dictionary and sEntiment Reasoner) simple negation, and (5) returns a weighted composite score. Furthermore, by adding Unicode-cleanup and Poppler-based PDF-to-text utilities, the package also broadens data-source coverage. In an illustrative example, on a corpus of 17,996 English tweets, EmojiSentR changed the predicted polarity in 20.7% of messages while adding only 769.4 ms of computation time per 1,000 posts. For instance, polarity reversals appeared in sarcastic laughter and uncertainty cues. The EmojiSentR package makes it simple and convenient to treat emojis as first-class sentiment carriers within R. Its modular design supports user-tunable weights, transparent lexicon updates, and forthcoming multilingual extensions, offering researchers a reproducible tool for analyzing emoji-rich text as well as PDF-based data sources.

Zitieren

APA

Hanson, L., Lahrim, E., & Tong, X. (2026). EmojiSentR: An R Package for Integrated Text and Emoji Sentiment Analysis. Journal of Behavioral Data Science, 6(1), 152-164. https://doi.org/10.35566/jbds/tonghl

Vancouver

Hanson L, Lahrim E, Tong X. EmojiSentR: An R Package for Integrated Text and Emoji Sentiment Analysis. Journal of Behavioral Data Science. 2026;6(1):152-164. doi:10.35566/jbds/tonghl

BibTeX
@article{Hanson2026,
  author = {Hanson, Logan and Lahrim, Elias and Tong, Xin},
  title = {EmojiSentR: An R Package for Integrated Text and Emoji Sentiment Analysis},
  journal = {Journal of Behavioral Data Science},
  year = {2026},
  volume = {6},
  number = {1},
  pages = {152--164},
  doi = {10.35566/jbds/tonghl}
}
Kostenlos starten →