envc — Electronic Nonverbal Cue Detection

Detect electronic nonverbal cues (eNVCs) in text-based communication. Implements a taxonomy grounded in nonverbal communication theory, covering two domains:

Domain	Subcategories
Kinesics	Stage directions (*hugs*), facial emoji, body/gesture emoji, emotion-conveying emoji, kaomoji
Paralinguistics	Vocalics (lol, ugh, sigh), intensity caps (AMAZING), intensity punctuation (!!!), elongation (loooove), alternating case (sPoNgEbOb), ellipsis (...)

From: Kumar & Jaidka (2026). Reading Between the Lines: How Electronic Nonverbal Cues Shape Emotion Decoding. ICWSM.

Please cite us! @inproceedings{kumarjaidka2026envc, author = {Taara Kumar and Kokil Jaidka}, title = {Reading Between the Lines: How Electronic Nonverbal Cues Shape Emotion Decoding}, booktitle = {Proceedings of the International AAAI Conference on Web and Social Media}, year = {2026} }

Python

Install

pip install envc
# or from source:
cd python && pip install -e ".[dev]"

Usage

from envc import detect_envc, annotate_df

# Single text
result = detect_envc("OMG I loooove this!! 😍😍")
print(result)
# {'stage_direction': False, 'emoji_faces': True, ..., 'elongation': True, 'any_envc': True}

# Count occurrences
from envc import detect_envc_counts
counts = detect_envc_counts("😍😍😍 wow!!")
print(counts)
# {'emoji_faces': 3, 'intensity_punctuation': 1, ...}

# Annotate a DataFrame
import pandas as pd
df = pd.DataFrame({"text": ["hello!!", "AMAZING 🎉", "meh"]})
annotated = annotate_df(df)
print(annotated[["text", "any_envc"]])

Run tests

cd python && pip install -e ".[dev]" && pytest

R

Install

# From GitHub (requires devtools):
devtools::install_github("kokiljaidka/envc", subdir = "R")

Usage

library(envc)

# Single text or vector
detect_envc("OMG I loooove this!! 😍😍")
detect_envc(c("hello!!", "AMAZING 🎉", "plain text"))

# Count occurrences
detect_envc_counts("😍😍😍 wow!!")

# Annotate a data frame
df <- data.frame(text = c("hello!!", "OMG 😍", "plain"))
annotate_envc(df)

Run tests

devtools::test("R")

Taxonomy → Regex → Output Mapping

Each detection category maps from nonverbal communication theory to a regex pattern to an output column in the results. This table mirrors Table 1 in the paper.

Domain	Subcategory	Regex Pattern	Example Matches	Output Column
Kinesics	Body / Touch	\*(hug|wave|frown|smile|clap)\*	hug, frowns	stage_direction
	Facial / Eye emoji	[\U0001F600-\U0001F64F]	😀 😂 😍	emoji_faces
	Body-part emoji	[\U0001F400-\U0001F4FF]	🙌 👏 💪	emoji_body
	Emotion-conveying emoji	[\U0001F490-\U0001F9E1]	❤️ ✨ 🎉	emoji_emotion
	Kaomoji	\([^\w\s]{1,3}[._][^\w\s]{1,3}\)	(^.^) (T_T)	kaomoji
Paralinguistics	Vocalics	\b(lol|yawn|ugh+|hmmm+)\b	lol, yawn, ughhh	vocalics
	Volume (caps)	\b[A-Z]{2,}\b	THIS, STOP	intensity_caps
	Volume (punctuation)	!!+ or \?\?+	!!!, ???	intensity_punctuation
	Pitch (elongation)	(\w)\1{2,}	soooo, noooo	elongation
	Pitch (alt. case)	\b([A-Z][a-z]){2,}\b	HiYa, LoL	alternating_case
	Ellipsis	\.{3,}|…	..., …	ellipsis

All functions return one boolean (or count) per column. The any_envc column is True if any category fires.

Citation

If you use this package, please cite:

@inproceedings{ahmed2026envc,
  title={Reading Between the Lines: How Electronic Nonverbal Cues Shape Emotion Decoding},
  author={Ahmed, Saif and Jaidka, Kokil},
  booktitle={Proceedings of the International AAAI Conference on Web and Social Media (ICWSM)},
  year={2026}
}

License

MIT