Word embedding for social sciences: an interdisciplinary survey

Machine learning models learn low-dimensional representations from complex high-dimensional data. Not only computer science but also social science has benefited from the advancement of these powerful tools. Within such tools, word embedding is one of the most popular methods in the literature. Howe...

Full description

Saved in:
Bibliographic Details
Published in:PeerJ. Computer science 2024-12, Vol.10, p.e2562
Main Authors: Matsui, Akira, Ferrara, Emilio
Format: Article
Language:English
Subjects:
Algorithms
Bias in machine learning
Computational linguistics
Computational social science
Computer science
Human acts
Human behavior
Language processing
Machine learning
Natural language interfaces
Surveys
Word embedding
word2vec
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Machine learning models learn low-dimensional representations from complex high-dimensional data. Not only computer science but also social science has benefited from the advancement of these powerful tools. Within such tools, word embedding is one of the most popular methods in the literature. However, we have no particular documentation of this emerging trend because this trend overlaps different social science fields. To well compile this fragmented knowledge, we survey recent studies that apply word embedding models to human behavior mining. Our taxonomy built on the surveyed article provides a concise but comprehensive overview of this emerging trend of intersection between computer science and social science and guides scholars who are going to navigate the use of word embedding algorithms in their voyage of social science research.
ISSN:2376-5992
2376-5992
DOI:10.7717/peerj-cs.2562