Repository logo
  • English
  • Deutsch
Log In
or
  1. Home
  2. HSG CRIS
  3. HSG Publications
  4. Why Do We Turn to Virtual Companions? A Text Mining Analysis of Replika Reviews
 
  • Details

Why Do We Turn to Virtual Companions? A Text Mining Analysis of Replika Reviews

Type
conference paper
Date Issued
2022
Author(s)
Siemon, Dominik
Strohmann, Timo
Khosrawi-Rad, Bijan
de Vreede, Triparna
Elshan, Edona  
Meyer, Michael
Research Team
IWI6
Abstract
Many people globally experience the feeling of loneliness and struggle with its consequences. A modern way
to deal with this loneliness and lack of companionship is to use empathetic and emotional conversational
agents. Often referred to as virtual companions, these agents can engage in human-like conversations with
their users and build relationships with them through modern artificial intelligence technologies. One
established service of such virtual companions is Replika, which we investigate in this study to explore what
users expect to gain from long-term interactions with virtual companions and what they tend to talk about
with them. Using a text mining approach and 119,831 reviews of the Replika service, we analyze users'
sentiments, emotions, and topics. Our results show that users interact with virtual companions to cope with
their loneliness and, especially, to address their mental well-being. Furthermore, Replika users have a joyful
and beneficial experience during long-term interaction with such virtual companions.
Language
English
HSG Classification
contribution to scientific community
HSG Profile Area
SoM - Business Innovation
Book title
Americas Conference on Information Systems (AMCIS)
Publisher place
Minneapolis, Minnesota, USA
Pages
10
Event Title
Americas Conference on Information Systems (AMCIS)
Event Location
Minneapolis, Minnesota, USA
Event Date
10-14 Aug 2022
URL
https://www.alexandria.unisg.ch/handle/20.500.14171/109536
Subject(s)

information managemen...

Division(s)

IWI - Institute of In...

Eprints ID
268159
File(s)
Loading...
Thumbnail Image

open.access

Name

JML_924.pdf

Size

343.07 KB

Format

Adobe PDF

Checksum (MD5)

6209ec1494376e1f70eedeb7b543c57e

here you can find instructions and news.

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science

  • Privacy policy
  • End User Agreement
  • Send Feedback