消費者購物參考之信任過程與評論內容因素分析-以Amazon.com為例

The Factor Analysis of Trust Process and Reviews’ Content to Customer as A Purchasing Reference: Case of Amazon.com

沈利芳(Li-Fang Shen)

指導教授 ： 邱淑芬

共同指導教授 ： 林真伊(Chen-Yi Lin)

國立臺中科技大學/資訊與流通學院/資訊管理系碩士班/碩士(2021年)

摘要

透過網路購物已經是現在購物的常態，賣家如果想要吸引更多的客源就必須要改進自己的賣場，包含產品、客服態度、發貨時間等，透過線上消費者評論了解自己的消費者便是很好的方法，但由於現在線上資訊量龐大，難以在短時間透過該資訊提出管理決策。而既有的研究僅使用單一調查方式，經常導致研究結果無法推估整個市場或確定該結果的應用場景。因此本研究以Amazom.com線上購物平台為例，期望利用問卷調查以及文字探勘的方式，結合信任過程因素分析以及評論內容因素分析，協助賣家有效率地了解消費者對於賣場的真實看法，進而改善自己的賣場、做出合宜的管理決策。
信任過程因素分析將評論功能利用歸納法分成評論標題、評論內容、產品星級評分和評論有用性這四個部分。針對曾經使用過Amazon.com進行購物的消費者，透過問卷調查的方式進行實驗，探討消費者在觀看評論時，評論的哪一個部分是信任評論的前置因素。前置因素分為基於信任轉移的信任網站、基於關聯理論的標題吸引力與標題簡潔度、基於來源可信度的內容可信度以及內容資訊性，與基於選擇性注意的產品星級評分和評論有用性。
評論內容因素分析則是使用Amazon.com線上購物網站中，電子消費產品的線上消費者評論，利用文字探勘的方式，結合預處理、術語頻率反文檔頻率(Term Frequency-Inverse Document Frequency, TF-IDF)和奇異值分析(Singular Value Decomposition, SVD)，使用分群的方式找出消費者經常討論的主題，及該主題中討論的方向為何。實驗將分為情境一單一產品測試探討同類別的兩件商品進行比較，和情境二單一類別測試使用同一類別產品的所有評論，並且比較兩種情境的使用結果及適用性。
本論文將提出兩大貢獻，信任過程因素分析中將根據線上消費者評論的功能使用狀況的，提出信任過程的模型架構。另外，在評論內容因素分析中，較少研究提出分析方法的應用場景，本論文將比較兩種場景的應用狀況，並提出兩者相異之處。

關鍵字

線上消費者評論 ； 信任評論過程 ； 評論內容因素

並列摘要

Shopping on the Internet is normal now. If sellers want to attract more customers, they must improve their stores, including products, service or delivery time, etc. It is a good way to understand their customers through online customer reviews, but the amount of online information is too huge now, and it is difficult to make management decisions through the information in a short time. Existing research is cannot estimate the overall market or determine the usage scenarios because their survey methods are relatively single. Therefore, this study uses the Amazom.com online shopping platform as an example. It is expected to use questionnaire surveys and text mining, combined with the factors analysis of trust process and the factors analysis of review’s content, to help sellers effectively understand customers’ true views on the store. Then improve their stores and make appropriate management decisions.
The factors analysis of trust process divides the review function into four parts: title, content, Product’s star rating and Usefulness vote. For customers who have used Amazon.com for shopping, we do the experiments through questionnaires, to explore which part of the review is the pre-factor when customers trust the review. The pre-factors are divided into ‘Trust in review site’ based on Transfer of Trust, Attractiveness of titles and Conciseness of titles base on Relevance Theory, Credibility of content and Informativeness of content base on Source Credibility, Product’s star ratings and Review’s helpful base on Selective Attention.
The factors analysis of review’s content is to use online customer reviews of electronic consumer products in Amazon.com, using text mining, the steps include preprocessing, Term Frequency-Inverse Document Frequency (TF-IDF) And Singular Value Decomposition (SVD). Use Clustering to find out topics that customers often discuss, and what is discussed in the topic. The experiment will be divided into scenario one, using a small amount of data to test and compare two products of the same category. Scenario two, using a large amount of data to test all reviews using the same category of products, and compare the results and applicability of the two scenarios.
This paper will propose two major contributions. In the factors analysis of trust process, we will propose the model framework of the trust process based on the function usage of online customer reviews. In addition, in the factors analysis of review’s content, there is less research on the application scenarios of the analysis methods. This paper will compare the application status of the two scenarios and propose the differences between the two.

並列關鍵字

Online customer reviews ； The process of trust in reviews ； The factors of review’s content

參考文獻

Banerjee, S., Chua, A. Y. (2019). Trust in online hotel reviews across review polarity and hotel category. Computers in Human Behavior, 90, 265-275.

Bansal, G., Zahedi, F. M., Gefen, D. (2016). Do context and personality matter? Trust and privacy concerns in disclosing private information online. Information Management, 53(1), 1-21.

Belanche, D., Casaló, L. V., Flavián, C., Schepers, J. (2014). Trust transfer in the continued usage of public e-services. Information Management, 51(6), 627-640.

Chen, J., Chen, J. E., Goh, K. Y., Xu, Y. C., Tan, B. C. (2014). When do sellers bifurcate from Electronic Multisided Platforms? The effects of customer demand, competitive intensity, and service differentiation. Information Management, 51(8), 972-983.

Cortinas, M., Cabeza, R., Chocarro, R., Villanueva, A. (2019). Attention to online channels across the path to purchase: An eye-tracking study. Electronic Commerce Research and Applications, 36.