Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Identifying Patterns in User Behavior in aMusic Streaming Service: A Cluster AnalysisApproach
KTH, School of Computer Science and Communication (CSC).
2013 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Logged user data has become a highly valued asset to many Internet based services with large user bases. Being able to draw insight from this data is considered a key to gaining competitive advantages for the companies behind the services. This study aims to identify patterns in the behavior of users when interacting with Spotify, a music streaming service, by studying automatically logged data. In the study, we examine several methods to perform such analyses using machine learning techniques. We identify six different types of behavior through k-means cluster analysis, each representing between 51.4% and 0.5% of all user sessions. We also identify five factors partly explaining the differences in behavior between different sessions. These are found through factor analysis and account for 39% of the variance in the data. Finally, we demonstrate how factors and clusters can be translated from numeric representations to linguistic interpretations.

Abstract [sv]

Loggad användardata har blivit en högt värderad tillgång för många Internetbaserade tjänster med en stor mängd användare. Att finna insikter från dessa data anses vara en nyckel till att vinna konkurrensfördelar för företagen bakom tjänsterna. Denna studie har som mål att identifiera mönster i beteendet hos användare av Spotify, en musikströmningstjänst, genom att studera loggad data. I studien utreds flera metoder för att göra denna typ av analys genom att använda maskininlärningstekniker. Vi identifierar sex olika typer av beteende genom k-means klusteranalys, där var och en representerar beteendet i mellan 51.4 % och 0.5 % av alla sessioner. Vi identifierar ocksa fem faktorer som förklarar en del av skillnaderna i beteende mellan användares olika sessioner. Dessa hittas genom faktoranalys, och förklarar tillsammans 39 % av variansen i studiens data. Till sist går vi igenom hur kluster och faktorer kan översättas från numeriska representationer till semantiska tolkningar.

Place, publisher, year, edition, pages
2013.
National Category
Computer Science
Identifiers
URN: urn:nbn:se:kth:diva-142038OAI: oai:DiVA.org:kth-142038DiVA: diva2:699672
Educational program
Master of Science in Engineering - Computer Science and Technology
Supervisors
Examiners
Available from: 2014-03-12 Created: 2014-02-28 Last updated: 2014-03-12Bibliographically approved

Open Access in DiVA

fulltext(1011 kB)760 downloads
File information
File name FULLTEXT01.pdfFile size 1011 kBChecksum SHA-512
9a20d8fa63b65bae17374abc0baf55912eb6b5c6c07c935e9a4763bff011e2c7ee7c3e13c83280520fa45150d5cb5672337cec40f405d1fcd2b055e051f000a5
Type fulltextMimetype application/pdf

By organisation
School of Computer Science and Communication (CSC)
Computer Science

Search outside of DiVA

GoogleGoogle Scholar
Total: 760 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 264 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf