交通行動服務使用者之套票購買行為分析－以高雄市MaaS系統為例

Analysis of User’s Purchasing Behavior of MaaS Packages－A case study of the MaaS System in Kaohsiung city

周翔淳(Shiang-Chung Chou)

指導教授 ： 盧宗成

國立交通大學/管理學院/運輸與物流管理學系/碩士(2020年)

摘要

交通行動服務（MaaS）為近年來被提出的一項新興概念，國內外許多城市已開始推動MaaS系統，了解民眾的套票購買及使用行為將有助於主管機關與業者推廣MaaS系統。本研究之目的在以決策樹建立一套會員方案續買預測模型，將高雄市MaaS系統的會員註冊資料、方案購買記錄以及電子票證搭乘記錄作為輸入資料，訓練與建立決策樹預測模型。經由初步資料處理與分析發現MaaS會員套票購買資料在各方案間的比例有顯著不平衡的現象，可能會影響決策樹模型的預測結果。為了改善此一問題，本研究提出以模擬方法增加樣本數的策略，先將此模擬方法與文獻中常用的SMOTE方法利用網路上公開的資料集進行測試與比較，發現模擬方法顯著優於SMOTE方法。接著本研究採用此方法處理MaaS會員資料，並且建構MaaS會員方案續買預測的決策樹模型，經由交叉驗證測試結果發現，由模擬方法增加樣本數的資料所建構的決策樹模式有不錯預測結果，精確率平均為84%，顯示此預測模型具備會員方案續買預測之能力。此外，本研究也針對決策樹模式的分支規則進行探討，發現除了使用者在各個運具的每月花費會影響續買行為，月份也是一個重要的考量因素。本研究成果可提供MaaS營運業者作為參考，藉由會員方案續買預測結果採取適當的行銷措施。

關鍵字

交通行動服務 ； 資料探勘 ； 決策樹 ； MaaS方案購買行為

並列摘要

Mobility as a service (MaaS) is an emerging concept in recent years, and it has been promoted in many cities around the world. Understanding MaaS users’ package-purchasing behavior will help the authorities and operators to promote MaaS. This study develops a package-purchasing prediction model using decision tree technique. The decision tree model is built and trained using the data of membership registration, package-purchasing records and iPASS card transit-ridership records of MaaS users in Kaohsiung. Through preliminary data processing and analysis, it is found that significant imbalance exists in MaaS users’ package-purchasing records for the different plans, and the imbalance may affect the prediction results of the model. To address this issue, the study proposes a strategy that uses simulation to generate addition sample. This method is first tested and compared with the SMOTE method commonly used in the literature by using datasets published on the Internet, and it is found that the simulation method is significantly better than the SMOTE method. Then the study uses the simulation method to balance the MaaS users’ packet-purchasing data, and constructs a decision tree model for MaaS users’ package-purchasing prediction. Through cross-validation test results, it is found that the model constructed by the oversampling data using the simulation method has good prediction results. The average accuracy is 84%, which shows that the prediction model has the ability to predict the users’ package-purchasing behavior. This study also discusses the branch variables in the decision tree model, and found that the user's monthly spending on each public transportation mode will affect the package-purchasing of MaaS users. Moreover, month is also an important variable. The results of this study can be used as a reference for MaaS operators to take appropriate actions on marketing based on the results of the users’ package-purchasing prediction.

並列關鍵字

Mobility as a service ； data mining ； decision tree ； purchasing behavior of MaaS packages

參考文獻

1. Gokasar, Ilgin, Kevser Simsek, and Kaan Ozbay. "Using big data of automated fare collection system for analysis and improvement of BRT-Bus rapid transit line in Istanbul." 94th Annual Meeting of the Transportation Research Board, Washington, DC. 2015.

2. Long, Ying, and Jean-Claude Thill. "Combining smart card data and household travel survey to analyze jobs–housing relationships in Beijing." Computers, Environment and Urban Systems 53 (2015): 19-35.

3. Lovelace, Robin, Ilan Fridman-Rojas, and Rob Long. "New tools of the trade? The potential and pitfalls of’Machine Learning’and’DAGs’ to model origin-destination data." GeoComputation 2017. Centre for Computational Geography, University of Leeds, 2017.

4. Meng, Chuishi, et al. "Travel purpose inference with GPS trajectories, POIs, and geo-tagged social media data." 2017 IEEE International Conference on Big Data (Big Data). IEEE, 2017.

5. Rodríguez, Jorge, et al. "Unsupervised Hierarchical Clustering Approach for Tourism Market Segmentation Based on Crowdsourced Mobile Phone Data." Sensors 18.9 (2018): 2972.