Predicting user demographics from music listening information

Abstract

Online activities such as social networking, online shopping, and consuming multi-media create digital traces, which are often analyzed and used to improve user experience and increase revenue, e. g., through better-fitting recommendations and more targeted marketing. Analyses of digital traces typically aim to find user traits such as age, gender, and nationality to derive common preferences. We investigate to which extent the music listening habits of users of the social music platform Last.fm can be used to predict their age, gender, and nationality. We propose a feature modeling approach building on Term Frequency-Inverse Document Frequency (TF-IDF) for artist listening information and artist tags combined with additionally extracted features. We show that we can substantially outperform a baseline majority voting approach and can compete with existing approaches. Further, regarding prediction accuracy vs. available listening data we show that even one single listening event per user is enough to outperform the baseline in all prediction tasks. We also compare the performance of our algorithm for different user groups and discuss possible prediction errors and how to mitigate them. We conclude that personal information can be derived from music listening information, which indeed can help better tailoring recommendations, as we illustrate with the use case of a music recommender system that can directly utilize the user attributes predicted by our algorithm to increase the quality of it's recommendations.

The relation of culture, socio-economics, and friendship to music preferences: A large-scale, cross-country study

“…According to a survey[84], these three musical metadata types are important factors for both creating playlist and managing music collections. In the existing literature, album [85, 86], artist [23, 60, 87–89], and genre listening [2, 23, 87, 90] are often used to measure users’ music preferences[8]. Therefore, incorporating album, artist, and genre is not only for providing a relatively comprehensive picture that reflects users’ music preferences, but also for the consideration of the setting of the current music retrieval and recommendation systems.…”

“…As some researchers stated, there is a tendency for people to be increasingly open to appreciating all genres [13, 59]. Nevertheless, recent research showed that even very little data about a listener’s artist and genre preferences is sufficient to predict his or her country of residence with high accuracy [60]. Therefore, it seems reasonable to assume that also genre can provide hints of the listener’s cultural background.…”

Characterizing DSM-5 and ICD-11 personality disorder features in psychiatric inpatients at scale using electronic health records

“…As anticipated based on studies using traditional personality measures, we observed an association between sociodemographic features and individual personality trait domains (Lynn and Martin, 1997; Kjelsås and Augestad, 2004). Demographic profiles are useful to predict certain behaviors (Krismayer et al ., 2019), but their relationship with dimensional traits is less studied (Al-Halabí et al ., 2010).…”

Using Word2Vec Recommendation for Improved Purchase Prediction

“…RS are used in many domains to filter relevant products for users in online platforms, including movie [20], ecommerce [4], and music [22]. There are many kinds of methods to develop RS with the most commonly used methods being Content-Based RS (CBRS) [40], Collaborative Filtering (CFRS) [42], and Hybrid-Based RS.…”