[Zurück]

Y. Deldjoo, M. Schedl, B. Hidasi, P. Knees:
"Multimedia Recommender Systems";
in: "Proceedings of the 12th ACM Conference on Recommender Systems", herausgegeben von: ACM; ACM, New York, NY, USA, 2018, ISBN: 978-1-4503-5901-6, S. 537 - 538.

This tutorial introduces multimedia recommender systems (MMRS), in particular, recommender systems that leverage multimedia content to recommend different media types. In contrast to the still most frequently adopted collaborative filtering approaches, we focus on content-based MMRS and on hybrids of collaborative filtering and content-based filtering. The target recommendation domains of the tutorial are movies, music and images. We present state-of-the-art approaches for multimedia feature extraction (text, audio, visual), including deep learning methods, and recommendation approaches tailored to the multimedia domain. Furthermore, by introducing common evaluation techniques, pointing to publicly available datasets specific to the multimedia domain, and discussing the grand challenges in MMRS research, this tutorial provides the audience with a profound introduction to MMRS and an inspiration to conduct further research.

deep learning, feature extraction, image recommender systems, multimedia recommender systems, music recommendation, video recommendation

http://dx.doi.org/10.1145/3240323.3241620

https://dl.acm.org/authorize?N660614

Projektleitung Peter Knees:
SmarterJam