旅遊業顧客不再訪重要因素分析之研究 - 以TripAdvisor飯店評論為例

A Study on Analyzing Important Factors of Customer Non-Revisit in Tourism – An Example of Hotel Reviews on TripAdvisor

曾恕慈(SHU-CIH TSENG)

指導教授 ： 陳隆昇

朝陽科技大學/資訊學院/資訊管理系/碩士(2020年)

摘要

隨著社群媒體的發展也改變了旅客對景點的訪問行為，對於旅遊業提升旅客再訪意圖是一直以來探討的研究重要議題，其目的是為了增加業者在市場上的競爭優勢。近年來的研究集中在再訪的旅客行為及因素探討，較少有研究探討旅客不再訪之相關議題，部分研究已顯示旅客若不再訪將可能帶來對業者在市場上不利的未來行為。為避免這些行為的發生，本研究將採用旅遊網站的線上文字評論進行資料探討，並整理過去潛在影響不再訪之因素，以TripAdvisor網站的飯店之旅客評論作為實驗案例，並利用決策樹(DT)、最小絕對壓縮挑選機制 (LASSO)、支持向量機之遞迴式特徵消除(SVM-RFE)特徵選取方法確認影響旅客不再訪飯店之因素，並使用支持向量機(SVM)與倒傳遞類神經網路(BPN)建構分類模型評估結果。最後根據分析結果將給予旅遊飯店服務者一些提升服務品質的建議，以有效減少旅客未來不再訪之意願。

關鍵字

社群媒體 ； 再訪意圖 ； 文字探勘 ； 特徵選取 ； 決策樹 ； 最小絕對壓縮挑選機制 ； 支持向量機之遞迴式特徵消除

並列摘要

With the rapid development of social media, it has changed the visitor’s behavior to tourist attractions. One of critial issues to increase the competitive strength in the tourism industry is to improve revisit rates for potential customers. Recently, related studies have focused on the revisit behaviors and factors, but just few researches mentioned why customers non-revisit. Some researches indicated non-revisit might bring the damage to tourism industry. To avoid these non-revisit situations happens, the study will adopt the online reviews on TripAdvisor to discover the critiacal factors of non-revisit. Three feature selection methods including Decision tree (DT), Least Absolute Shrinkage and Selection Operator (LASSO) and Support Vector Machine Recursive Feature Elimination (SVM-RFE) will be employed to select important factors. Then Support Vector Machine (SVM) and Backpropagation Network (BPN) will be testify the effectiveness of selected factor set. According to the results of analysis, the study can provide some suggestions to improve the service quality and to decrease the non-revisit willingness.

並列關鍵字

Social media ； Revisit intention ； Text mining ； Feature Selection ； Decision tree ； LASSO ； SVM-RFE

參考文獻

[1] 陳隆昇、林孟儒(2017)，確認影響手機遊戲App內消費之關鍵因素，碩士論文，朝陽科技大學資訊管理系，臺中。

[2] A. F. Chao and H. L. Yang(2018), “Using Chinese radical parts for sentiment analysis and domain-dependent seed set extraction,” Computer Speech and Language, vol. 47, pp. 194-213.

[3] A. Ganzaroli, I. De Noni and P. van Baalen(2017), “Vicious advice: Analyzing the impact of TripAdvisor on the quality of restaurants as part of the cultural heritage of Venice,” Tourism Management, vol. 61, pp. 501-510.

[4] A. J. Kim and K. K. Johnson(2016), “Power of consumers using social media: Examining the influences of brand-related user-generated content on Facebook,” Computers in Human Behavior, vol. 58, pp. 98-108.

[5] A. Kumar and T. Gilovich(2016), “To do or to have, now or later? The preferred consumption profiles of material and experiential purchases,” Journal of Consumer Psychology, vol. 26, no. 2, pp. 169-178.