About me
I am an assistant professor at Polytechnic University of Bari (Politecnico di Bari), Italy, affiliated with the Information Systems Laboratory (SisInf Lab), led by Prof. Tommaso di Noia, the head of the group. My research focuses on: (i) designing multimodal information seeking systems (recommendation, question answering and information retrieval systems), and (ii) designing and evaluating machine learning models to understand their robustness, fairness, generalizability, and interpretability. Recent and ongoing projects consider the fairness of recommender systems, adversarial machine learning for the security of recommender systems, federated learning for user privacy, multimedia recommendation systems, and multimodal conversational systems.
I defended my Ph.D. dissertation, with "con lode" (highest distinction in Italy) in July 2018! I have received two bachelor's degrees one a B.Sc. in Electrical Engineering from the University of Guilan, and the other a B.A. degree in English, Linguistics from the University of Gothenburg, Sweden. I completed my M.Sc. in Electrical Engineering at the Chalmers University of Technology, Sweden and obtained a Ph.D. degree in Computer Science from Polytechnic University of Milan (Politecnico di Milano), Italy. During my Ph.D., I was a visiting researcher at Johannes Kepler University (JKU) Linz, Austria for a period of 6 months.
I am an active member of the recommender systems (RS) community and regularly publish/present tutorials at the major venues, such as RecSys, SIGIR, WSDM, and top-tier journals such as ACM Computing surveys (CSUR), UMUAI, IEEE TKDE, and ACM TIST. In the year 2020, I had two surveys accepted at ACM CSUR centered on the topics outlined above. I also had the honor to contribute to two book chapters at the 3rd Edition of the RS handbook. I have been involved in organizing ACM Recommender System Challenges as part of the RecSys conference through 2017-2020 (as a chair or advisor). I have also organized two tasks at the MediaEval benchmarking event in 2018 and 2019. .
NEW: Our comprehensive survey on "Multimedia Recommender Systems" accepted to ACM Computing Surveys is now availiable online. Domains studied in this work include: fashion, audio (music, sounds), video (movie, user-generated videos), news, social media, food, e-commerce, tourism, among others: https://bit.ly/3jTlJY5
NEW: Our new work "A Flexible Framework for Evaluating User and Item Fairness in Recommender Systems" accepted to UMUAI, the main journal for personalization research. It extends our RMSE@RecSys'19 paper in several dimensions: large-scale experiment, theoretical support and extensive literature review! PrePrint is now availible https://bit.ly/33KSYYr
NEW: The second version of our recent extensive literature review "A Survey on Adversarial Recommender Systems: from Attack/Defense Strategies to Generative Adversarial Networks" under review at ACM computing Surveys has now a prePrint version available online. https://bit.ly/2A3TlAo
NEW: A new dataset "Session-Based Hotel Recommendations Dataset" co-authered with Trivago has been accepted to ACM TIST. The link to dataset is provided within the paper. https://bit.ly/2OcguVm
news
July-2020 Our paper on recommender systems fairness evaluation is accepted to UMUAI!July-2020 Our tutorial on adversarial machine learning in RecSys is accepted to RecSys'20.July-2020 Our paper session-based hotel recommendations dataset done in collobration with Trivago is accepted at ACM TIST.July-2020 Our paper federated learning is accepted to Italian journal of Intelligenza Artificiale.June-2020 Our comprehensive literature review about recommender systems leveraging multimedia content is accepted to ACM Computing Surveys.May-2020 Our literature review on adversarial machine learning in recommender systems has a preprint version.Jan-2020 One full paper accepted to SIGIR 2020.Dec-2019 One full paper accepted to ESWC 2020.Oct-2019 Our tutorial on adversarial machine learning in RecSys accepted to WSDM'20.Oct-2019 A US patent accepted.July-2019 Our paper on federated learning accepted to AIIA'19.Jan-2019 Our paper accepted to UMUAI.
Current Research Topics
- Recommender systems and personalization (RecSys)
- Multimedia Recommender Systems (MM-RecSys)
- Fairness in Recommender Systems (Fair-RecSys)
- Advarsarial Machine Learning in Recommender Systems (Security-RecSys)
- Federated and Privacy-Aware Recommender Systems (Privacy-RecSys)
Seleted Publications
- (MM-RecSys) Yashar Deldjoo, Markus Schedl, Paolo Cremonesi, Gabriella Pasi. "Recommender Systems Leveraging Multimedia Content.", ACM Computing Surveys 2020.
- (Fair-RecSys) Yashar Deldjoo, Vito Walter Anelli, Hamed Zamani, Alejandro Bellogin, Tommaso Di Noia. "A Flexible Framework for Evaluating User and Item Fairness in Recommender Systems", UMUAI 2020.
- (Security-RecSys) Yashar Deldjoo, Tommaso Di Noia, Felice Antonio Merra. "How Dataset Characteristics Affect the Robustness of Collaborative Recommendation Models", SIGIR 2020.
- (MM-RecSys) Yashar Deldjoo, Maurizio Ferrari Dacrema, Mihai Gabriel Constantin, Hamid Eghbal-Zadeh, Stefano Cereda, Markus Schedl, Bogdan Ionescu, Paolo Cremonesi. "Movie Genome: Alleviating New Item Cold Start in Movie Recommendation", UMUAI 2019.
- (MM-RecSys) Yashar Deldjoo, Mihai Gabriel Constantin, Hamid Eghbal-zadeh, Markus Schedl, Bogdan Ionescu, Paolo Cremonesi."Audio-Visual Encoding of Multimedia Content to Enhance Movie Recommendations", RecSys 2018.
- (Privacy-RecSys) Vito Walter Anelli, Yashar Deldjoo, Tommaso Di Noia, Antonio Ferrara. "Towards Effective Device-Aware Federated Learning", AIIA 2019.
Academic Services
- Organizer: ACM RecSys Challenge 2019 (session-based hotel recommendation/Trivago), ACM RecSys Challenge 2017 (Job Recommendation/Xing), MediaEval 2019 (MovieRec and NEWSReel), MediaEval 2018 (MovieRec)
- PC Member: SIGIR'20, ACM MM'20, ECIR'20, UMAP'20, RecSys'19, ACM MM'19, UMAP'19, ECIR'19, MMSys'19
- Journal Reviewer: ACM Computing Surveys (CSUR), Journal of User Modeling and User-Adapted Interaction (UMUAI), Elsevier Expert Systems with Applications, IEEE Access