Personalizing POI Descriptions with LLMs: An Exploratory Study of User Perceptions

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Points of Interest (POI) descriptions on travel platforms are typically generic, presenting the same content to all users regardless of individual interests and travel preferences. While personalization in tourism systems usually focuses on selecting and ranking places, the way POIs are described also influences how users evaluate them. This work examines whether large language models (LLMs) can personalize POI descriptions while preserving perceived information quality. In a within-subjects user study, 51 participants evaluated original and LLM-adapted descriptions across 10 POIs without knowing which version had been adapted. Exploratory descriptive results indicate higher mean ratings for clarity and perceived trustworthiness for LLM-adapted descriptions, while perceived significance remains comparable. These findings are positioned as initial exploratory evidence for description-level personalization in Human-Centred AI and tourism HCI, and supplementary materials are provided to support transparency and replication.