This study aims to restructure a location-aware audio guide mobile application designed for urban walking tours. Traditional points of interest-based geofencing, which triggers automatic guide delivery as users approach specific locations, struggle to provide continuous and consistent storytelling in areas with limited notable spots, thereby diminishing tourist experiences. To address this challenge, we propose a hierarchical geofencing framework that forms the basis for seamless audio guide experiences through scale-based feature switching and the definition of story serialization rules. In addition, this study proposes geofence-to-conversation techniques utilizing text-to-speech engines and large language models to dynamically adapt guide document resources to dynamic tourists’ movements. A demonstration conducted in a historic urban park area highlighted that the guide generation time in both English and Japanese guide modes is significantly shorter than playback duration, confirming technical feasibility for seamless regional storytelling. Furthermore, we define metrics such as starting time errors, ending time errors, undelivered time to evaluate the real-time performance of location-aware audio guide applications. The experimental results demonstrate effective strategies for geofence configuration and operations, enhancing user experiences in our generative audio tours. This intelligent guide approach, designed for complex urban environments, is expected to enrich tourism and foster regional learning.

A circular geofence (CGF), a virtual boundary defined by a center point and a radius, plays a significant role in various location-based services. Our paper discusses the need for automatic generation of geofences due to the difficulties in creating adaptable settings for user movement and GPS sensor errors regarding one of the most typical mobile tour guide applications - a service helping city tourists discover nearby points of interest. Currently, human-created fixed CGFs sharing the same center point of points of interest have been generally adopted thanks to their intuitive nature and simplicity. However, our algorithm for automatic parameter determination based on tourists' GPS logs enables us to introduce flexible CGFs that can be placed independently of the position of actual geographic features. Our experimental study using a GPS dataset indicates that this new style of geofences could produce more relevant notifications for a maximum number of tourists.

This study proposes a method for generating geofences driven by GPS trajectory data to realize scalable point-of-interest (POI) notifiers, encouraging walking tourists to discover new local spots. The case study revealed that manual geofence settings degrade the location relevance and user coverage—key objectives of POI notifiers—and hinder the scalability and reliability of services. The formalization presented computationally equips geofence designers with practical solutions through two implementations based on prior GPS trajectory logs: (1) a multiobjective genetic algorithm that suggests cost-effective geofences by providing trade-off visualizations and (2) a user coverage-penalized genetic algorithm that determines an optimal geofence based on the designers’ expectations. The feasibility and stability of the proposed implementations were tested in areas with varying tourist flow patterns. A comparative survey among manual settings, settings incorporating a reliability simulation, and data-driven settings demonstrates significant performance improvements for geofence services.

Location-based storytelling is a key strategy in smart tourism, enabling immersive engagement through narrative content triggered by user movement. This study proposes a data-driven framework for designing enclosure- and viewpoint-geofences that align with storytelling modes, leveraging GPS horizontal-accuracy clustering and motion-sensor noise filtering to detect meaningful spatial engagement zones. We introduce a lightweight decision-tree classifier using cluster duration and motion variability features to distinguish valid indoor stays from transient or noise-induced clusters. In a field experiment with 12 participants in Akita City, our method achieved 94 percent classification accuracy and a 0.96 F₁-score under leave-one-out cross-validation. Furthermore, our qualitative comparisons imply the geofence identifications can outperform baseline techniques such as HDBSCAN and stay point detection. The results demonstrate the practical potential of the proposed approach for context-sensitive geofencing in urban tourism. This framework advances autonomous, adaptive geofencing for enriched tourist experiences.

This paper proposes a model-less feedback system driven by tourist tracking data that are automatically collected through mobile applications to visualize the gap between geomedia recommendations and the actual routes selected by tourists. High-frequency GPS data essentially make it difficult to interpret the semantic importance of hot spots and the presence of street-level features on a density map. Our mobile collaborative framework reorganizes tourist trajectories. This processing comprises (1) extracting the location of the user-generated content (UGC) recording, (2) abstracting the locations where tourists stay, (3) discarding locations where users remain stationary, and (4) simplifying the remaining points of location. Then, our heatmapping system visualizes heatmaps for hot streets, UGC-oriented hot spots, and indoor-oriented hot spots. According to our experimental study, this method can generate a trajectory that is more adaptable for hot street visualization than the raw trajectory and a simplified trajectory according to its geometry. This paper extends our previous work at the 2022 IEEE International Conference on Big Data, providing deeper discussions on application for local tourism. The framework allows us to derive insights for the development of guide content from mobile sensor data.

This study presents a geofence-based service architecture for city-wide audio augmented reality, tailored for the era of large language models. Traditional geofencing mechanisms, which monitor user entry to geofences, struggle to provide continuous storytelling in areas with few points of interest, degrading the audio tour experiences for pedestrians. Our proposed geofencing architecture consistently incorporates complex and multilayered city features, enabling seamless audio tour experiences. Furthermore, this paper introduces prompt engineering for generating entertaining guide scripts for large language models, that is, the geofence-to-conversation technique. The mobile application developed for the actual field demonstrates the feasibility of our proposed architecture and highlights future challenges in enhancing users’ interaction with a city.