Skip to content

Dataset · betterup

Tabular data from stranger video calls in CANDOR corpus

betterup/candor-tabular

Millions of records from a multimodal corpus of natural conversations between strangers via video call.

What's inside

This part of the large-scale CANDOR corpus from BetterUp Labs contains tabular data (1 to 10 million rows) corresponding to video recordings of unscripted, spontaneous conversations between pairs of unfamiliar adult volunteers from the US. The recordings were made via video call in 2020 with participants from the Prolific platform and published as part of a study in Science Advances. Metadata specifies tabular, text, and video modalities in English.

What you can do with it

  • Analyzing natural dialogue dynamics, turns, and participant behavior in conversational systems
  • Training models to predict engagement and conversational shifts based on tabular features
Languages
English
Size
1M–10M rows
Downloads this month
6.5K
Likes
30
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("betterup/candor-tabular")