Wisdom of the Crowd, Without the Crowd: A Socratic LLM for Asynchronous Deliberation on Perspectivist Data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Khadar, Malik, Runningen, Daniel, Tang, Julia, Chancellor, Stevie, Kaur, Harmanpreet
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916896485408768
author Khadar, Malik
Runningen, Daniel
Tang, Julia
Chancellor, Stevie
Kaur, Harmanpreet
author_facet Khadar, Malik
Runningen, Daniel
Tang, Julia
Chancellor, Stevie
Kaur, Harmanpreet
contents Data annotation underpins the success of modern AI, but the aggregation of crowd-collected datasets can harm the preservation of diverse perspectives in data. Difficult and ambiguous tasks cannot easily be collapsed into unitary labels. Prior work has shown that deliberation and discussion improve data quality and preserve diverse perspectives -- however, synchronous deliberation through crowdsourcing platforms is time-intensive and costly. In this work, we create a Socratic dialog system using Large Language Models (LLMs) to act as a deliberation partner in place of other crowdworkers. Against a benchmark of synchronous deliberation on two tasks (Sarcasm and Relation detection), our Socratic LLM encouraged participants to consider alternate annotation perspectives, update their labels as needed (with higher confidence), and resulted in higher annotation accuracy (for the Relation task where ground truth is available). Qualitative findings show that our agent's Socratic approach was effective at encouraging reasoned arguments from our participants, and that the intervention was well-received. Our methodology lays the groundwork for building scalable systems that preserve individual perspectives in generating more representative datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2508_09911
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Wisdom of the Crowd, Without the Crowd: A Socratic LLM for Asynchronous Deliberation on Perspectivist Data
Khadar, Malik
Runningen, Daniel
Tang, Julia
Chancellor, Stevie
Kaur, Harmanpreet
Human-Computer Interaction
Data annotation underpins the success of modern AI, but the aggregation of crowd-collected datasets can harm the preservation of diverse perspectives in data. Difficult and ambiguous tasks cannot easily be collapsed into unitary labels. Prior work has shown that deliberation and discussion improve data quality and preserve diverse perspectives -- however, synchronous deliberation through crowdsourcing platforms is time-intensive and costly. In this work, we create a Socratic dialog system using Large Language Models (LLMs) to act as a deliberation partner in place of other crowdworkers. Against a benchmark of synchronous deliberation on two tasks (Sarcasm and Relation detection), our Socratic LLM encouraged participants to consider alternate annotation perspectives, update their labels as needed (with higher confidence), and resulted in higher annotation accuracy (for the Relation task where ground truth is available). Qualitative findings show that our agent's Socratic approach was effective at encouraging reasoned arguments from our participants, and that the intervention was well-received. Our methodology lays the groundwork for building scalable systems that preserve individual perspectives in generating more representative datasets.
title Wisdom of the Crowd, Without the Crowd: A Socratic LLM for Asynchronous Deliberation on Perspectivist Data
topic Human-Computer Interaction
url https://arxiv.org/abs/2508.09911