A survey of Human-AI Interaction for Personalized

The Hong Kong University of Science and Technology (Guangzhou)

Data Science and Analytics Thrust

PhD Qualifying Examination

By Ms. Chuyi LI

Abstract

Personalized itinerary recommendation has emerged as a key application of AI within the

travel and tourism industry, leveraging user-specific data to create customized travel plans.

The rise of Human-AI interaction has significantly enhanced the ability of recommendation

systems to provide nuanced, user-centered travel suggestions. While these AI techniques have

shown promise in other fields, their integration into personalized itinerary planning is still

developing, with most existing systems relying on a combination of rule-based methods and

recommendation algorithms. Effective integration of LLMs, KGs, and RAG enables systems

to interpret complex geospatial and user preference data, adapt in real time, and offer highly

relevant recommendations for each individual. This survey presents a comprehensive review

of Human-AI interaction techniques in personalized itinerary recommendation, organized by

key technological approaches and methods. First, we define the core concepts and explore

various personalized recommendation tasks relevant to itinerary planning. We then categorize

and examine the major methods used in these tasks, including content-based, collaborative

filtering, and hybrid recommendation systems. Additionally, we review state-of-the-art Human

AI collaboration techniques that support itinerary recommendation, detailing the roles of LLMs,

KGs, and RAG in enhancing geospatial information processing and user interaction. Finally, we

identify several promising research directions, aiming to guide future work in this evolving field

and unlock new potential for AI-driven, user-centric itinerary planning systems.

PQE Committee

Chair of Committee: Prof. Qiong LUO

Prime Supervisor: Prof. Lei LI

Co-Supervisor: Prof. Wei ZENG

Examiner: Prof. Zeyi WEN