[1J3-OS-10a] OS-10

[1J3-OS-10a-04] Proposal for an Automated Tool to Generate Improved Communication Strategies through the Analysis of Gaps between Tourist Site Information and Visitor Perceptions

YUKIO OHSAWA¹, KAIRA SEKIGUCHI¹, 〇RIKO KIMURA¹ (1. Univ. of Tokyo)

Keywords:Tourism , Large Language Model, Gap Analysis

Information about tourist destinations is usually disseminated by the residents of the area. However, there often exists a gap between this information and the attractions felt by visitors. Understanding the factors that cause this gap and finding solutions are important. How to narrow or eliminate this gap in the dissemination of information about tourist destinations, and specific methods of improvement are still not clear. The purpose of this study is to identify the differences in attractions between official tourism information and visitor reviews, and to propose a method to clarify these differences. For the relevant data, similarity is calculated using a Transformer model. The differences from the group of official information that are similar are explained by the Transformer model. Specifically, the vectors of the documents are normalized, and similarity is calculated using the dot product. The top 10 documents with high similarity are selected, and sentences that generate the commonalities and differences with specific reviews are created. These sentences are evaluated by experts in the tourism industry. A tool for improving the dissemination of tourism information was developed and its accuracy was evaluated. This made it possible to identify areas for improvement in information dissemination at the target tourist destination.