Doctoral course - Recommendation System

Extent 3 ECTS

Abstract

The course builds on the expanding need to provide a holistic and global approach for tackling the recommendation issue, which is highly present in several PhD topics ranging from pure engineering field to pure mathematical fields passing through social, business, medical and artificial intelligence fields, where the choice and recommendation doctrine is highly present in the research agenda.

Therefore, this course provides basics of recommendation system, highlighting the three types of recommender systems (content-based, collaborative filtering and hybrid), elaborating how we can use item-content and user’s preferences according to profile item-ranking to generate item recommendation, while it includes several case studies to fit tailored needs from various disciplines. Moreover, the course describes various approaches to elicit user’s preference and how to integrate this into the recommendation system. Both mathematical models of key recommendation systems and algorithmic construction will be described.

The course is organized as a three-day intensive course, one day for self-exercises for projects done by students, and one extra day for feedback and ways forward.

Day 1: Basic of recommendation systems (3 h lectures in morning) + 2h lecture in afternoon + 2h exercise & tutorial

Day 2: Case studies: (3h lectures + discussions in morning) + 2h discussion in afternoon + 1h Initiation to python implementation of recommender systems (group activities)

Day 3: Selected case studies on recommender system design (3h Python exercises in morning) + 3h discussion on case studies and initiation to group projects exercises to be done in Day 4 in pairs.

Day 4: Self-exercises (assistant can be called upon whenever needed). Students submit their projects at the end of the day.

Day 5: Morning (3h): Assessment of projects, feedback to students and discussion of possible collaboration opportunities

Assessment: Group project.

Short biography of Invited lecturer

Sajad Ahmadian received his B.S. degree in computer engineering from Razi University, Kermanshah, Iran, in 2011, and the M.Sc. degree in computer engineering, artificial intelligence from University of Kurdistan, Sanandaj, Iran, in 2014 and a PhD Ph.D. degree in computer engineering, artificial intelligence from University of Zanjan, Zanjan, Iran, in 2018. He conducted a part of his Ph.D. research work in the laboratory of complex networks, RMIT University, Melbourne, Australia, form February 2018 to July 2018. He is currently the acting Dean at the Faculty of Information Technology, Kermanshah University of Technology, Kermanshah, Iran. His research interests include recommender systems, social networks analysis, data mining, Healthcare informatics, and machine learning where he led several research projects. In 2022, he visited University of Oulu as part of DigiHealth Visiting programme.