Details

Defining and predicting users' listening mode on streaming platforms from interaction metadata
ID Bevec, Matej (Author), ID Pesek, Matevž (Mentor) More about this mentor... This link opens in a new window, ID Tkalčič, Marko (Comentor)

.pdfPDF - Presentation file, Download (685,93 KB)
MD5: 5CB272EC2BA4C214B36B10B70524D07D

Abstract
As streaming services have made music ubiquitous, users increasingly expect their listening experiences to adapt to their current situation. While contextual recommendation systems aim to meet this need, they typically rely on external context and overlook internal factors such as emotional state. Consequently, although generally seen as valuable, they often fail to align with user expectations. Motivated by user-study-centered research, which shows that people engage with music through distinct interaction paradigms, we propose that listening intent can be inferred directly from behavior. Using a Spotify dataset, we cluster listening sessions based on interpretable interaction features to identify listening modes and predict them early in a session. Our analysis reveals three distinct modes. Passive listening involves minimal interaction. In active exploration, users navigate around the platform, manually exploring tracks, while active refinement sees targeted skipping to fine-tune existing collections. Passive listening may be related to music discovery, whereas active refinement may reflect live curation for a particular mood, though additional data is needed to confirm this. We show that listening modes can be predicted with promising accuracy, improving as more interactions are observed. Contrary to some research, we find that while background listening is frequent, active refinement is even more prevalent, and that skipping may sometimes reflect curation rather than dissatisfaction. These findings can help streaming platforms anticipate user expectations and adapt their interfaces accordingly. Beyond system design, our work provides data-driven insights relevant to user-centered fields, such as human-computer interaction and multimedia, and illuminates how users engage with music in the streaming era.

Language:English
Keywords:music recommendation systems, context-aware recommendation, user modeling, streaming platforms, music discovery, background listening
Work type:Master's thesis/paper
Typology:2.09 - Master's Thesis
Organization:FRI - Faculty of Computer and Information Science
Year:2025
PID:20.500.12556/RUL-175341 This link opens in a new window
COBISS.SI-ID:255968259 This link opens in a new window
Publication date in RUL:24.10.2025
Views:177
Downloads:36
Metadata:XML DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:Slovenian
Title:Opredelitev in predikcija uporabnikovega načina poslušanja na pretočnih platformah iz metapodatkov o interakciji
Abstract:
Odkar pretočne storitve ponujajo stalen dostop do glasbe, uporabniki pričakujejo, da si lahko poslušalsko izkušnjo prilagodijo svoji trenutni situaciji. Trenutno se ta izziv naslavlja s kontekstualnimi priporočilniki, ki se opirajo na eksterni kontekst, nimajo pa dostopa do notranjih dejavnikov, kot je razpoloženje. Posledično so ti sistemi, čeprav koristni, pogosto dojeti kot neskladni z uporabniškimi pričakovanji. Da bi naslovili te omejitve, se opiramo na pretekle raziskave na osnovi uporabniških študij, ki kažejo, da uporabniki z glasbo komunicirajo preko različnih interakcijskih paradigm, ter preizkusimo, če je uporabnikov namen mogoče sklepati neposredno iz vedenjskih vzorcev. Z gručenjem poslušalskih sej na platformi Spotify zato definiramo načine poslušanja in jih skušamo napovedati že zgodaj v seji. Naša analiza razkriva tri različne načine. Za pasivno poslušanje je značilna minimalna interakcija s storitvijo. Pri aktivnem raziskovanju se uporabniki angažirano pomikajo po platformi in raziskujejo skladbe, medtem ko pri aktivnem prilagajanju s preskoki filtrirajo obstoječe zbirke. Raziskovanje je morda povezano z odkrivanjem nove glasbe, prilagajanje pa s kuracijo specifičnega razpoloženja, čeprav so za potrditev te domneve potrebni dodatni podatki. Poleg tega se izkaže, da je načine poslušanja mogoče napovedati z obetavno natančnostjo, ki se izboljšuje tekom seje. V nasprotju z nekaterimi raziskavami ugotavljamo, da je sicer poslušanje v ozadju pogosto, a je aktivno prilagajanje še pogostejše, in tudi, da preskakovanje skladb v določenih primerih ne odraža nezadovoljstva, temveč kuracijo. Ta spoznanja lahko pretočnim storitvam omogočijo bolje predvideti pričakovanja uporabnikov in temu primerno prilagoditi vmesnike. Poleg tega naše delo ponuja na podatkih osnovan vpogled v koncepte iz področij, kot sta interakcija človek-računalnik in multimedija, ter razkriva, kako uporabniki v dobi pretočnih storitev komunicirajo z glasbo.

Keywords:priporočilni sistemi za glasbo, kontekstualna priporočila, modeliranje uporabnikov, pretočne platforme, odkrivanje glasbe, poslušanje v ozadju

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back