從部落格探索客家旅遊目的地意象之研究——自然語言處理的方法與應用

An Investigation on the Images of Hakka Culture Tourism Destinations by Using Natural Language Processing

部落格興起讓旅遊經驗能在網路中被廣泛地傳遞與分享。在旅遊經驗敘事中遊客會表達個人對旅遊目的地發展的喜好及感受。針對一特定旅遊目的地蒐集網路中所有個人旅遊經驗進行分析，可以從遊客觀點了解旅遊目的地之觀光價值及整體意象發展，有助於地方觀光經營者掌握遊客需求與態度，做為經營規劃重要參考。然龐大網路口碑資訊短時間要有效分析處理則需要文字探勘的技術。本研究利用線上文字探勘系統，從自然語言處理概念，提出一個適用網路口碑的分析處理程序。以台灣北、中、南、東主要客家旅遊目的地為例，說明如何透過半自動化文字處理技術，探索遊客對旅遊目的地認知與態度，希望能提供另一種分析旅遊資訊的方法，充實觀光旅遊研究的應用範疇。

The emergence of blogs has enabled personal tourism experiences to be extensively propagated and shared on the Internet. Through the narration of tourism experiences, tourists’ express personal preferences and feelings toward local tourism developments and tourism destinations. By collecting and analyzing personal tourism experiences of a specific tourism destination, the corresponding tourism values and overall image developments can be observed from the perspectives of tourists, providing a crucial reference for the local tourism managers to make decisions based on tourists’ needs and attitudes. However, effectively analyzing abundant word-of-mouth information on the Internet within a limited time requires the use of text mining. This study adopts an online text mining system and, from the concept of natural language processing, proposes a processing procedure suitable for analyzing online word-of-mouth information. Using Hakka culture tourism destinations from northern, central, southern, and eastern Taiwan as examples, we explain how to use the semi-automated text processing technique for determining the perceptions and attitudes of tourists toward tourism destinations. This study provides an alternative method for analyzing tourism information and broadens the scope of applications in tourism studies.