以集群分析探討電商平台資料特性：以蝦皮購物I公司為例

CCUR > College of Business > Master Program of Electronic Commerce > thesis & dissertation > Item 987654321/53292

Please use this identifier to cite or link to this item: https://irlib.pccu.edu.tw/handle/987654321/53292

Title:	以集群分析探討電商平台資料特性：以蝦皮購物I公司為例 Exploring the data characteristics of e-commerce platforms through cluster analysis: Case study on I company in Shopee
Authors:	彭晴
Contributors:	電子商務碩士學位學程
Keywords:	電子商務大數據資料探勘集群分析新冠肺炎 E-commerce Big Data Data Mining Cluster Analysis COVID-19
Date:	2024
Issue Date:	2024-03-21 13:02:22 (UTC+8)
Abstract:	科技的進步促使網路逐漸普及化，也塑造出一個不同於以往的新興消費模式—電子商務。2020年新冠肺炎在全球各地蔓延逐漸造成大流行，人們因為疫情不得不改變原本的生活模式，也因此將電子商務的發展推上了另一個新高峰，並且被看好其未來市場需求仍會持續成長。電子商務為企業開拓了更多可能性，因此若懂得掌握電子商務之特性並善加利用勢必能讓企業保持優勢地位。本研究以蝦皮購物商城商家為個案研究對象，並蒐集2020年7月至2023年2月之間共計973筆的每日交易資料作為研究數據，欲探討疫情間消費者之線上消費行為。為了找出蝦皮購物商城商家疫情期間每日交易資料的異同之處，本研究使用資料探勘中的 K-means集群分析法對其進行分析，研究結果得出共6個不同特徵之集群，依序為表現良好群、表現低落群、表現穩定群、表現有潛力群、表現優異群、表現特殊群，最後並依據各集群結果，於結論部分提供建議予品牌企業端作為擬訂行銷經營策略之參考，同時提出未來研究方向，後續研究者可以此進行不同角度的深入探討。 The advance of technology has led to the gradual popularization of the Internet and shaped a new emerging business model, namely e-commerce. The COVID-19 pandemic, which spread globally in 2020, has forced people to change their lifestyles and consequently propelled the development of e-commerce to new heights. Moreover, the future market demand for e-commerce is expected to continue growing. E-commerce opens up more possibilities for businesses and understanding its characteristics and leveraging them effectively can undoubtedly help enterprises maintain a competitive advantage. This study focuses on the Shopee sellers as the case study subjects. A total of 973 daily transaction data from July 2020 to February 2023 were collected as research data to investigate consumer online shopping behavior during the pandemic. To identify the similarities and differences in daily transaction data of Shopee sellers during the pandemic. This study utilizes K-Means clustering for data mining to conduct the analysis. The research findings reveal six distinct clusters with different characteristics: High-Performance Cluster, Low-Performance Cluster, Stable Performance Cluster, Potential Performance Cluster, Outstanding Performance Cluster, and Special Performance Cluster. Finally, based on the results of each cluster, recommendations are provided in the conclusion for brand enterprises to formulate marketing and business strategies. Additionally, future research directions are proposed to encourage further exploration from different perspectives by subsequent researchers.
Appears in Collections:	[Master Program of Electronic Commerce] thesis & dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	102	View/Open

View Licence

Loading...