Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques

A novel method of exploring linguistic networks is introduced by mapping word-adjacency networks to time series and applying multifractal analysis techniques. This approach captures the complex structural patterns of language by encoding network properties—such as clustering coefficients and node de...

Full description

Saved in:
Bibliographic Details
Main Authors: Jakub Dec, Michał Dolina, Stanisław Drożdż, Robert Kluszczyński, Jarosław Kwapień, Tomasz Stanisz
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Entropy
Subjects:
Online Access:https://www.mdpi.com/1099-4300/27/4/356
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849713844466745344
author Jakub Dec
Michał Dolina
Stanisław Drożdż
Robert Kluszczyński
Jarosław Kwapień
Tomasz Stanisz
author_facet Jakub Dec
Michał Dolina
Stanisław Drożdż
Robert Kluszczyński
Jarosław Kwapień
Tomasz Stanisz
author_sort Jakub Dec
collection DOAJ
description A novel method of exploring linguistic networks is introduced by mapping word-adjacency networks to time series and applying multifractal analysis techniques. This approach captures the complex structural patterns of language by encoding network properties—such as clustering coefficients and node degrees—into temporal sequences. Using Alice’s Adventures in Wonderland by Lewis Carroll as a case study, both traditional word-adjacency networks and extended versions that incorporate punctuation are examined. The results indicate that the time series derived from clustering coefficients, when following the natural reading order, exhibits multifractal characteristics, revealing inherent complexity in textual organization. Statistical validation confirms that observed multifractal properties arise from genuine correlations rather than from spurious effects. Extending this analysis by taking into account punctuation equally with words, however, changes the nature of the global scaling to a more convolved form that is not describable by a uniform multifractal. An analogous analysis based on the node degrees does not show such rich behaviors, however. These findings reveal a new perspective for quantitative linguistics and network science, providing a deeper understanding of the interplay between text structure and complex systems.
format Article
id doaj-art-efe72dbbab6e4c9a8bd82c69f2887f3f
institution DOAJ
issn 1099-4300
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Entropy
spelling doaj-art-efe72dbbab6e4c9a8bd82c69f2887f3f2025-08-20T03:13:51ZengMDPI AGEntropy1099-43002025-03-0127435610.3390/e27040356Exploring Word-Adjacency Networks with Multifractal Time Series Analysis TechniquesJakub Dec0Michał Dolina1Stanisław Drożdż2Robert Kluszczyński3Jarosław Kwapień4Tomasz Stanisz5Faculty of Computer Science and Telecommunications, Cracow University of Technology, 31-155 Kraków, PolandFaculty of Computer Science and Telecommunications, Cracow University of Technology, 31-155 Kraków, PolandFaculty of Computer Science and Telecommunications, Cracow University of Technology, 31-155 Kraków, PolandComplex Systems Theory Department, Institute of Nuclear Physics, Polish Academy of Sciences, 31-342 Kraków, PolandComplex Systems Theory Department, Institute of Nuclear Physics, Polish Academy of Sciences, 31-342 Kraków, PolandComplex Systems Theory Department, Institute of Nuclear Physics, Polish Academy of Sciences, 31-342 Kraków, PolandA novel method of exploring linguistic networks is introduced by mapping word-adjacency networks to time series and applying multifractal analysis techniques. This approach captures the complex structural patterns of language by encoding network properties—such as clustering coefficients and node degrees—into temporal sequences. Using Alice’s Adventures in Wonderland by Lewis Carroll as a case study, both traditional word-adjacency networks and extended versions that incorporate punctuation are examined. The results indicate that the time series derived from clustering coefficients, when following the natural reading order, exhibits multifractal characteristics, revealing inherent complexity in textual organization. Statistical validation confirms that observed multifractal properties arise from genuine correlations rather than from spurious effects. Extending this analysis by taking into account punctuation equally with words, however, changes the nature of the global scaling to a more convolved form that is not describable by a uniform multifractal. An analogous analysis based on the node degrees does not show such rich behaviors, however. These findings reveal a new perspective for quantitative linguistics and network science, providing a deeper understanding of the interplay between text structure and complex systems.https://www.mdpi.com/1099-4300/27/4/356quantitative linguisticscomplex networksvertex observablesclustring coefficienttime seriesmultiscale correlations
spellingShingle Jakub Dec
Michał Dolina
Stanisław Drożdż
Robert Kluszczyński
Jarosław Kwapień
Tomasz Stanisz
Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques
Entropy
quantitative linguistics
complex networks
vertex observables
clustring coefficient
time series
multiscale correlations
title Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques
title_full Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques
title_fullStr Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques
title_full_unstemmed Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques
title_short Exploring Word-Adjacency Networks with Multifractal Time Series Analysis Techniques
title_sort exploring word adjacency networks with multifractal time series analysis techniques
topic quantitative linguistics
complex networks
vertex observables
clustring coefficient
time series
multiscale correlations
url https://www.mdpi.com/1099-4300/27/4/356
work_keys_str_mv AT jakubdec exploringwordadjacencynetworkswithmultifractaltimeseriesanalysistechniques
AT michałdolina exploringwordadjacencynetworkswithmultifractaltimeseriesanalysistechniques
AT stanisławdrozdz exploringwordadjacencynetworkswithmultifractaltimeseriesanalysistechniques
AT robertkluszczynski exploringwordadjacencynetworkswithmultifractaltimeseriesanalysistechniques
AT jarosławkwapien exploringwordadjacencynetworkswithmultifractaltimeseriesanalysistechniques
AT tomaszstanisz exploringwordadjacencynetworkswithmultifractaltimeseriesanalysistechniques