Language Use and Susceptibility in Online Conversation

Lu Xiao, Qiyi Wu, Sucheta Soundarajan, Jinfen Li

Research output: Chapter in Book/Entry/PoemConference contribution


Prior study indicates that persuasion attempts in small and private online communications can be very influential. Yet, most existing online persuasion studies focus on large-scale and open online discussions. In this study, we investigated whether and how one’s language use indicates their susceptibility to persuasion in one-to-one synchronous online text-based chats. We analyzed 815 one-to-one online discussions by 321 pairs. Our results show that discussions in which one or both participants change their views tend to have more positive emotions, more affective processes, and more impersonal pronouns. Additionally, individuals who did not change their minds tend to focus more on problem solving whereas those who changed their minds focus more on the relationship building. Our findings imply the potential of using surface level linguistic features in predicting the persuasion outcome in a one-to-one online discussion, shedding light on the development of persuasive dialogue system which is on the rise.

Original languageEnglish (US)
Title of host publicationIntelligent Computing - Proceedings of the 2022 Computing Conference
EditorsKohei Arai
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages13
ISBN (Print)9783031104633
StatePublished - 2022
EventComputing Conference, 2022 - Virtual, Online
Duration: Jul 14 2022Jul 15 2022

Publication series

NameLecture Notes in Networks and Systems
Volume507 LNNS
ISSN (Print)2367-3370
ISSN (Electronic)2367-3389


ConferenceComputing Conference, 2022
CityVirtual, Online


  • One-to-One discussion
  • Online persuasion
  • Susceptibility

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Signal Processing
  • Computer Networks and Communications


Dive into the research topics of 'Language Use and Susceptibility in Online Conversation'. Together they form a unique fingerprint.

Cite this