MSc Thesis Proposal by:
Mobin Ali Momin
Date: 21-Jan-2026
Time: 1:30PM
Location: Online (MS Teams)
Meeting ID: 231 381 809 281 91
Abstract:
Planning a one-day itinerary usually is challenging because it requires selecting and ordering places that match preferences while staying feasible under real-world constraints such as limited time, travel time between locations, and (in theme parks) time-varying queue times. The challenge becomes harder for groups, where members have different interests and a single “best” itinerary can unfairly favor some users over others.
In this proposal, we first review existing work related to group recommendation and itinerary planning to identify current limitations and open challenges. We then introduce and describe our proposed approach, G-AIR (Group-Aware Intention-based Itinerary Recommendation System), a framework that generates an ordered, time-feasible group itinerary by combining graph-based POI representation learning with sequential intention modeling and fairness-aware preference aggregation.
Our aim is to generate feasible and balanced group itineraries by jointly modeling individual preferences, their temporal dynamics, and group-level aggregation. Finally, we discuss the planned evaluation methodology, including datasets, experimental settings, and metrics to assess recommendation quality, feasibility, and fairness.
We expect that our approach improves overall group satisfaction while providing more balanced and fair recommendations compared to single-user and non-fair group baselines.
Thesis Committee:
Internal Reader: Dr. Saeed Sameet
Internal Reader: Dr. Muhammad Asaduzzaman
Advisor: Dr. Pooya Moradian Zadeh