Idea of using word2vec in recommendation and advertizing is illustrated in an article by Chris McCormick.
Goal of the project is to build an effective word2vec-based music recommendation system, by translating user playlist into a word2vec-embedding, which utilizes the context of a track (other tracks) to map similar songs close to each other [1]. T
Donwload the playlists dataset of #nowplaying project [2] from http://dbis-nowplaying.uibk.ac.at/.
pip install numpy pandas sklearn gensim matplotlib
[1] https://arxiv.org/pdf/1804.04212.pdf Caselles-Dupré, Hugo, Florian Lesaint, and Jimena Royo-Letelier. "Word2Vec applied to Recommendation: Hyperparameters Matter." arXiv preprint arXiv:1804.04212 (2018).
[2] https://www.evazangerle.at/wp-content/uploads/2017/06/somera15.pdf Pichl, Martin, Eva Zangerle, and Günther Specht. "Towards a context-aware music recommendation approach: What is hidden in the playlist name?." Data Mining Workshop (ICDMW), 2015 IEEE International Conference on. IEEE, 2015.