Alessandro B. Melchiorre

alessandro.b.mel ((at)) gmail.com

prof_pic.jpg

Hola!

I am a Ph.D. student at the Institute of Computational Perception and at the Multimedia Mining and Search Group at the Johannes Kepler University Linz, Austria.

I studied Engineering in Computer Science in my Bachelor at Università degli Studi di Napoli Federico II and in my Master at Sapienza - Università di Roma where I both graduated with full marks.

My main interests revolve around the topics of recommender system algorithms, explainability in AI, and bias & fairness. I am particularly enthusiastic about developing interpretable models and explainability methods for recommender systems, especially in the music domain. I am also interested in investigating the relationships between users’ characteristics and music preference and consumption.

news

Sep 11, 2022 The Ars Electronica Festival is over. We had such a blast at our exhibit Black Holes of Popularity! In case you missed you can check the teaser video here.
Jul 11, 2022 Our research paper “ProtoMF: Prototype-based Matrix Factorization for Effective and Explainable Recommendations” has been accepted at RecSys 2022! Looking forward to travel to Seattle this September.
Apr 14, 2022 The article “Explainability in Music Recommender Systems” will appear in the the Summer 2022 Issue of the Ai Magazine. The article is the result of a very enjoyable and productive collaboration with my supervisor Markus Schedl and Darius Afchar, Romain Hennequin, Elena V. Epure, and Manuel Moussallam from Deezer Research.
Apr 13, 2022 The demo paper “EmoMTB: Emotion-aware Music Tower Blocks” has been accepted to the ACM International Conference on Multimedia Retrieval. I will travel to Newark, NJ, USA to present our demo to the guests of the conference.
Feb 24, 2022 The project proposal “Black Holes of Popularity” for the Linz Ars Electronica Festival 2022 has been accepted and will exhibited at this year’s festival! I will lead the project as PI for seven months together with my colleague Oleg Lesota and my supervisor Markus Schedl.

selected publications

  1. ProtoMF: Prototype-based Matrix Factorization for Effective and Explainable Recommendations
    Melchiorre, Alessandro B.Rekabsaz, NavidGanhör, Christian,  and Schedl, Markus
    In Proceedings of the 16th ACM Conference on Recommender Systems (RecSys) 2022
  2. Explainability in Music Recommender Systems
    Afchar, Darius*,  Melchiorre, Alessandro B.*Schedl, Markus, Hennequin, Romain, Epure, Elena V.,  and Moussallam, Manuel
    AI Magazine 2022
  3. Investigating Gender Fairness of Recommendation Algorithms in the Music Domain
    Information Processing & Management (IPM) 2021
  4. Personality Bias of Music Recommendation Algorithms
    Melchiorre, Alessandro B.Zangerle, Eva,  and Schedl, Markus
    In Fourteenth ACM Conference on Recommender Systems (RecSys) 2020