應用文字探勘推論旅客使用機場自助服務行李托運之關鍵要素

Abstract

近年來，全球主要國家力推機場智慧化及自動化。為了提升旅客及航空公司於機場之營運效率，大部份機場已經設置自助報到專區，更積極推動使用自助行李托運櫃檯專區等自助服務科技，以求優化旅客報到流程、提昇通關速率、強化服務品質、滿足機場智能應用需求。然而，旅客對於自助行李托運的接受度和使用頻率，仍有待提昇。接受度與使用頻率過低，勢必影響機場及航空公司未來是否持續建置自助行李托運櫃台。本議題至關重要，但少有研究探討。因此，本研究擬探討影響旅客使用機場自助行李托運系統之關鍵因素。首先，透過文字探勘技術，蒐集「PPT論壇」相關貼文等，經斷詞後，提取詞彙，再以隱含狄利克雷分佈 (Latent Dirichelet Allocation，LDA) 擷取主題，作為影響旅客使用機場自助行李托運服務之關鍵要素。之後，以技術接受模式為基礎，將關鍵要素歸入理論模型各構面中，並使用偏最小平方法結構方程 (Partial Least Squares Structural Equation Modeling，PLS-SEM)驗證技術接受模式各路徑是否顯著。依據分析結果，感覺有用、感覺易用均為顯著影響旅客使用自助行李托運服務之關鍵要素，本研究結果可供機場或航空公司於規劃設置自助行李托運服務、訂定推廣策略之用，分析架構也可作探討影響消費者接受其他新興運輸科技意圖之用。

In the era of continuous technological advancement, airports around the world are fully promoting airport intelligence and automation services. In order to raise the efficiency of the airport, the self-service check-in kiosks area have been generally set up. Now, airports are actively building self-service technologies such as self-service bag drop and automatic boarding gates. All in order to optimize passenger check-in and customs clearance rates, and strengthen service quality to enhance airport intelligence services. However, passengers' acceptance of self-service bag drop and frequency of use will inevitably affect whether airports and airlines’ implement a large number of self-service bag drop counters in the future. This research intends to explore the analysis of the key factors of passengers using airport self-service bag drop, and its purpose is to understand the behavioral intention and acceptance of passengers. Use text mining technology to search for related articles and other information in “PPT BBS”. After using JIEBA for word segmentation and extracting keywords, the part-of-speech tagging can make emotional words more quickly found in the analysis process. And performing topic modeling to build themes, the key factors that affect travelers' use of the self-service bag drop are obtained. After that, the results of the previous analysis are inferred as hypotheses. The text of the proposal is quantified in the form of topic distribution. The statistical method of Partial Least Squares Structural Equation Modeling (PLS-SEM) is used for data analysis. The key elements that affect passengers' use of self-service bag drop services at the airport are obtained, and whether the impact relationship is significant and the relative weight between the evaluation criteria are verified. The research results of this study can be used as a reference for airports or airlines to consider when planning to set up self-service bag drop.