Investigating the Relationship Between Diversity in Music Consumption Behavior and Cultural Dimensions: A Cross-Country Analysis

Diversity in recommendation lists or sets has shown to be an important feature in recommender systems as it can counteract on negative effects such as choice difficulty and choice overload. However, how much diversity a recommendation list needs to provide is not clearly defined. By analyzing music listening behavior of listeners in 47 countries, we show that diversity needs may be cultural dependent. For our analyses, we exploited a Last.fm dataset containing almost 1.1 billion single listening events. We investigated several diversity measures to identify how users in different countries apply music diversity to their listening behavior. By analyzing 53,309 Last.fm users, we found distinct diversity behavior related to several cultural dimensions of Hofstede. We show with our results that different diversity needs exist between cultures, and should be taken into account when applying diversity to a recommendation list.

[1]  Bruce Ferwerda,et al.  Fusing Social Media Cues: Personality Prediction from Twitter and Instagram , 2016, WWW.

[2]  Fabian Flöck,et al.  Mining cross-cultural relations from Wikipedia: A study of 31 European food cultures , 2014, WebSci.

[3]  Mor Naaman,et al.  Understanding Musical Diversity via Online Social Media , 2015, ICWSM.

[4]  Bruce Ferwerda,et al.  Personality Traits and the Relationship with (Non-) Disclosure Behavior on Facebook , 2016, WWW.

[5]  Alfons Trompenaars,et al.  Riding the Waves of Culture: Understanding Diversity in Global Business , 1993 .

[6]  Markus Schedl,et al.  The LFM-1b Dataset for Music Retrieval and Recommendation , 2016, ICMR.

[7]  P. Dorfman,et al.  Leadership and Organizations: The GLOBE Study of 62 Societies , 2004 .

[8]  Bruce Ferwerda,et al.  Using Instagram Picture Features to Predict Users' Personality , 2016, MMM.

[9]  Bruce Ferwerda,et al.  Personality Correlates for Digital Concert Program Notes , 2015, UMAP.

[10]  P. Todd,et al.  Can There Ever Be Too Many Options? A Meta-Analytic Review of Choice Overload , 2010 .

[11]  Fabio Celli,et al.  Automatic Personality and Interaction Style Recognition from Facebook Profile Pictures , 2014, ACM Multimedia.

[12]  Jennifer Golbeck,et al.  Predicting Personality from Twitter , 2011, 2011 IEEE Third Int'l Conference on Privacy, Security, Risk and Trust and 2011 IEEE Third Int'l Conference on Social Computing.

[13]  Shinobu Kitayama,et al.  Cultural Shaping of Self, Emotion, and Well‐Being: How Does It Work? , 2007 .

[14]  Mark P. Graus,et al.  Using latent features diversification to reduce choice difficulty in recommendation lists , 2011 .

[15]  Bruce Ferwerda,et al.  Personality Traits Predict Music Taxonomy Preferences , 2015, CHI Extended Abstracts.

[16]  Bruce Ferwerda,et al.  Enhancing Music Recommender Systems with Personality Information and Emotional States: A Proposal , 2014, UMAP Workshops.

[17]  Gregory J. Park,et al.  Automatic personality assessment through social media language. , 2015, Journal of personality and social psychology.

[18]  Mark P. Graus,et al.  Understanding choice overload in recommender systems , 2010, RecSys '10.

[19]  Bruce Ferwerda,et al.  Predicting Personality Traits with Instagram Pictures , 2015, EMPIRE@RecSys.

[20]  Bruce Ferwerda,et al.  Exploring Music Diversity Needs Across Countries , 2016, UMAP.

[21]  Y. Aharoni.  Cultures and Organizations: Software of the Mind , 1992 .

[22]  Bruce Ferwerda,et al.  Using Social Media Mining for Estimating Theory of Planned Behaviour Parameters , 2014, UMAP Workshops.

[23]  Juliane M. Stopfer,et al.  Facebook Profiles Reflect Actual Personality, Not Self-Idealization , 2010, Psychological science.

[24]  Daniele Quercia,et al.  Our Twitter Profiles, Our Selves: Predicting Personality with Twitter , 2011, 2011 IEEE Third Int'l Conference on Privacy, Security, Risk and Trust and 2011 IEEE Third Int'l Conference on Social Computing.