Skip to content
Thesis Info
- LABS ID
- 00682
- Thesis Title
- Musical Serendipity – Designing for Contextual Music Recommendation and Discovery
- Author
- Pirkka Åman
- E-mail
- pirkkaaman AT gmail.com
- 2nd Author
- 3rd Author
- Degree
- Doctor of Arts
- Year
- 2018
- Number of Pages
- 192
- University
- Aalto University
- Thesis Supervisor
- Lassi A. Liikkanen
- Supervisor e-mail
- lassial AT gmail.com
- Other Supervisor(s)
- Language(s) of Thesis
- English
- Department / Discipline
- School of Arts, Design and Architecture / Media / New Media
- Copyright Ownership
- Languages Familiar to Author
- English, Finnish
- URL where full thesis can be found
- www.dropbox.com/sh/59jydb28z67s6rz/AADoHBZRpNFTurO61CgpyWS8a?dl=0
- Keywords
- Music recommendation, music discovery, urban computing, context-awareness, context-aware music services, serendipity, urban augmentation, location-awareness, mobile computing, pervasive computing, ubiquitous computing, field studies.
- Abstract: 200-500 words
- Online media services with recommendation facilities provide the users with a possibility to tailor the recommended content such as books, TV series and music to the users’ personal preferences. Online music recommendation services are a subset of personalizable media services. As music preferences vary greatly across music listening situations, information about the user’s situation, in other words, context information, is sometimes involved in modern music recommendation services.
In this thesis, I suggest new ways of including context information, mainly location, to interactive music recommendations by presenting various concepts and prototypes. In the articles, I presented two prototypes, Sounds of Helsinki (Article II) and OUTMedia (Article IV), as well as a platform for several context-aware music service concepts (Article III). Furthermore, two articles reviewed the existing music services for their explanations and transparency (Article I) and the ways they involved context factors in interacting with music recommendation and discovery tasks (Article V).
The underlying argument and a starting point for the thesis was that by involving context factors, ultimately cultural diversity could be fostered. In an ideal case, adding context to music recommendations would lead to recommendations that offer more non-mainstream music than channels such as the playlist radio or playlists of new releases of online music services. That would lead to better chances for serendipitous discoveries, and, ultimately, given that the user base would be large enough, would promote cultural diversity as well.
While the results show that the users indeed experienced serendipity in many ways, in the light of the results it can not be proven that context-aware music recommendations necessarily lead to cultural diversity. In addition, the results can not be generalized to all context-aware music recommendation cases. Instead, design implications are given to help designers and researchers of future systems to build rewarding and enjoyable context-aware content services, especially to enrich urban environments. The implications include Supporting open meaning-making through combinations of different media content and places; Visual and interactive UI elements that communicate the system logic or explain why a recommendation was made; Positive restrictions, such as allowing the content to be available only when the user is near a certain location or within a defined time window; Supporting serendipity can be approached in many ways, for example by combining music with an activity, a location, certain time or an identity, which may result in serendipitous discoveries of not only music but the cultural layers of urban environments as well.