Publikation:

Visual Pattern Analytics for Event Sequences

Lade...
Vorschaubild

Dateien

Jentner_2-yd3pqfc7gh5p1.pdf
Jentner_2-yd3pqfc7gh5p1.pdfGröße: 15.73 MBDownloads: 96

Datum

2023

Herausgeber:innen

Kontakt

ISSN der Zeitschrift

Electronic ISSN

ISBN

Bibliografische Daten

Verlag

Schriftenreihe

Auflagebezeichnung

DOI (zitierfähiger Link)
ArXiv-ID

Internationale Patentnummer

Angaben zur Forschungsförderung

Projekt

Open Access-Veröffentlichung
Open Access Green
Core Facility der Universität Konstanz

Gesperrt bis

Titel in einer weiteren Sprache

Publikationstyp
Dissertation
Publikationsstatus
Published

Erschienen in

Zusammenfassung

Pattern mining plays an essential role in unsupervised machine learning as it allows the clustering of structured data without requiring distance measures and purely relying on the definition of containment. Because it is unsupervised, it is predestined for exploratory analysis, and visual analytics offers a holistic perspective thoroughly involving the data, task, and especially the user in the decision-making process of designing tools for exploratory analysis. Pattern mining can easily generate millions of patterns since the search spaces are exponential. Additionally, the structures are often large and complex, which thwarts sense-making efforts by the user. This dissertation explains how visual analytics can be leveraged to allow the effective exploration of sequentially structured data using pattern mining algorithms. The first focus is on interesting measures, a concept known from data mining that should quantify interestingness. Because interestingness is subjective and heavily depends on the task and the user, this work argues for understanding interestingness measures as features that quantify different properties of the patterns and the clusters they represent. It further presents an alternative taxonomy of available features that can be used in pattern mining and discusses their importance and limitations. Secondly, this work surveys visualization techniques for structured data patterns, including their features, and highlights the differences between structured data as the input for the mining and the patterns themselves. Furthermore, it discusses the limitations of the visualization techniques, especially concerning scalability and the number of features. Finally, well-known visual analytics concepts such as interactive visualizations, progressive visual analytics, or concepts from visual text analytics are being transferred for pattern mining and the exploration of patterns. It is explained and discussed how these concepts can be exploited and implemented to mitigate the effects of the exponential search spaces and the complexity of the patterns to ease the user’s burden during the exploration process. Even though this work focuses on event sequences and sequential patterns, all aspects can be transferred onto different data structures and pattern mining algorithms. Therefore, this dissertation provides a foundation for the exploratory analysis of structured data using pattern mining with countless possible extensions to inspire future research.

Zusammenfassung in einer weiteren Sprache

Fachgebiet (DDC)
004 Informatik

Schlagwörter

Konferenz

Rezension
undefined / . - undefined, undefined

Forschungsvorhaben

Organisationseinheiten

Zeitschriftenheft

Zugehörige Datensätze in KOPS

Zitieren

ISO 690JENTNER, Wolfgang, 2023. Visual Pattern Analytics for Event Sequences [Dissertation]. Konstanz: University of Konstanz
BibTex
@phdthesis{Jentner2023Visua-69007,
  year={2023},
  title={Visual Pattern Analytics for Event Sequences},
  author={Jentner, Wolfgang},
  address={Konstanz},
  school={Universität Konstanz}
}
RDF
<rdf:RDF
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:bibo="http://purl.org/ontology/bibo/"
    xmlns:dspace="http://digital-repositories.org/ontologies/dspace/0.1.0#"
    xmlns:foaf="http://xmlns.com/foaf/0.1/"
    xmlns:void="http://rdfs.org/ns/void#"
    xmlns:xsd="http://www.w3.org/2001/XMLSchema#" > 
  <rdf:Description rdf:about="https://kops.uni-konstanz.de/server/rdf/resource/123456789/69007">
    <dcterms:issued>2023</dcterms:issued>
    <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/36"/>
    <dcterms:rights rdf:resource="https://rightsstatements.org/page/InC/1.0/"/>
    <dcterms:available rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2024-01-11T11:25:30Z</dcterms:available>
    <dcterms:title>Visual Pattern Analytics for Event Sequences</dcterms:title>
    <foaf:homepage rdf:resource="http://localhost:8080/"/>
    <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/36"/>
    <dc:date rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2024-01-11T11:25:30Z</dc:date>
    <dc:contributor>Jentner, Wolfgang</dc:contributor>
    <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/>
    <dcterms:abstract>Pattern mining plays an essential role in unsupervised machine learning as it allows the clustering of structured data without requiring distance measures and purely relying on the definition of containment. Because it is unsupervised, it is predestined for exploratory analysis, and visual analytics offers a holistic perspective thoroughly involving the data, task, and especially the user in the decision-making process of designing tools for exploratory analysis. Pattern mining can easily generate millions of patterns since the search spaces are exponential. Additionally, the structures are often large and complex, which thwarts sense-making efforts by the user. This dissertation explains how visual analytics can be leveraged to allow the effective exploration of sequentially structured data using pattern mining algorithms. The first focus is on interesting measures, a concept known from data mining that should quantify interestingness. Because interestingness is subjective and heavily depends on the task and the user, this work argues for understanding interestingness measures as features that quantify different properties of the patterns and the clusters they represent. It further presents an alternative taxonomy of available features that can be used in pattern mining and discusses their importance and limitations. Secondly, this work surveys visualization techniques for structured data patterns, including their features, and highlights the differences between structured data as the input for the mining and the patterns themselves. Furthermore, it discusses the limitations of the visualization techniques, especially concerning scalability and the number of features. Finally, well-known visual analytics concepts such as interactive 
visualizations, progressive visual analytics, or concepts from visual text analytics are being transferred for pattern mining and the exploration of patterns. It is explained and discussed how these concepts can be exploited and implemented to mitigate the effects of the exponential search spaces and the complexity of the patterns to ease the user’s burden during the exploration process. Even though this work focuses on event sequences and sequential patterns, all aspects can be transferred onto different data structures and pattern mining algorithms. Therefore, this dissertation provides a foundation for the exploratory analysis of structured data using pattern mining with countless possible extensions to inspire future research.</dcterms:abstract>
    <dc:language>eng</dc:language>
    <dc:creator>Jentner, Wolfgang</dc:creator>
    <bibo:uri rdf:resource="https://kops.uni-konstanz.de/handle/123456789/69007"/>
    <dcterms:hasPart rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/69007/4/Jentner_2-yd3pqfc7gh5p1.pdf"/>
    <dc:rights>terms-of-use</dc:rights>
    <dspace:hasBitstream rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/69007/4/Jentner_2-yd3pqfc7gh5p1.pdf"/>
  </rdf:Description>
</rdf:RDF>

Interner Vermerk

xmlui.Submission.submit.DescribeStep.inputForms.label.kops_note_fromSubmitter

Kontakt
URL der Originalveröffentl.

Prüfdatum der URL

Prüfungsdatum der Dissertation

April 27, 2023
Hochschulschriftenvermerk
Konstanz, Univ., Diss., 2023
Finanzierungsart

Kommentar zur Publikation

Allianzlizenz
Corresponding Authors der Uni Konstanz vorhanden
Internationale Co-Autor:innen
Universitätsbibliographie
Ja
Begutachtet
Diese Publikation teilen