#Nowplaying on #Spotify: Leveraging Spotify Information on Twitter for Artist Recommendations
The rise of the web enabled new distribution channels like online stores and streaming platforms, offering a vast amount of different products. For helping customers finding products according to their taste on those platforms, recommender systems play an
- PDF / 397,681 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 78 Downloads / 196 Views
Abstract. The rise of the web enabled new distribution channels like online stores and streaming platforms, offering a vast amount of different products. For helping customers finding products according to their taste on those platforms, recommender systems play an important role. Besides focusing on the computation of the recommendations itself, in literature the problem of a lack of data appropriate for research is discussed. In order to overcome this problem, we present a music recommendation system exploiting a dataset containing listening histories of users, who posted what they are listening to at the moment on the microblogging platform Twitter. As this dataset is updated daily, we propose a genetic algorithm, which allows the recommender system to adopt its input parameters to the extended dataset. In the evaluation part of this work, we benchmark the presented recommender system against two baseline approaches. We show that the performance of our proposed recommender is promising and clearly outperforms the baseline. Keywords: Music recommender systems Social media · Twitter
1
·
Collaborative filtering
·
Introduction
The way how consumers access music has changed in recent years due to the rise of the web. Nowadays, consumers have the possibility to access a huge amount of different music using various devices and services. An example for such new services are music streaming platforms. The distribution and especially the inventory costs of such platforms are lower than the costs of traditional channels, i.e., brick and mortar stores. Due to this development, an increased amount of more diverse music is available, as additionally to bestsellers also niche music can be offered with low additional costs [1]. Besides commercial vendors like Spotify1 or Pandora2 , there are also open platforms like SoundCloud3 or Promo DJ4 , 1 2 3 4
http://www.spotify.com http://www.pandora.com http://soundcloud.com http://promodj.com
c Springer International Publishing Switzerland 2015 F. Daniel and O. Diaz (Eds.): ICWE 2015 Workshops, LNCS 9396, pp. 163–174, 2015. DOI: 10.1007/978-3-319-24800-4 14
164
M. Pichl et al.
where users can upload and publish their own creations, increasing the diversity of available music. In contradiction to traditional channels like the radio, these new channels allow consumers to freely choose tracks they listen to and to create own playlists. One drawback of this freedom of choice is, that it is difficult for the customers to identify new tracks they like and want to listen to in this sheer amount of artists and tracks. As explained in Section 2, recommender systems can be seen as method to deal with information overload and a type of searching for information [7]. Thus, implementing a recommender system helps users discovering new tracks according to their taste, which increases user satisfaction. As most data corpora in this field are owned by private companies, e.g., Spotify or Pandora mentioned above, this data is not or only publicly available limited. Thus, we propose to utilize Twit
Data Loading...
