以約略集理論輔助網路商店提昇電子交易品質之探討

Upgrade Electronic Transaction Quality of On-line Shopping by Using the Rough Set Theory

胡維萍(Wei-Ping Hu) ； 陳雅玲(Ya-Ling Chen)

《資訊管理展望》 12卷2期 (2010/12) Pp. 201-219

https://doi.org/10.29715/JIMCSA.201012.0005

摘要

對於網路商店而言，提供最佳的品質給顧客，並能即時回應顧客需求與期望，制定提昇品質的新決策，是增加消費者滿意度與忠誠的主因，本研究根據Bauer et al.(2006)所提：「電子交易品質」模型中交易品質量測因素，架構整個品質資訊系統，並利用決策規則下情況屬性映射，找出輔助網路商店提昇電子交易品質決策，與探討相關性能。本文將電子交易品質測量因素，設計成「約略集」(Rough set)理論中的「宇集合」(Universe set)，而其相對應的「網站的功能性與設計」、「娛樂」、「交易過程」、「可靠性」、「回應性」五個指標設為「屬性集合」，經由隨機抽樣進行問卷調查，可建構出「不可辨識關係表」，並完成「約略分類」(Rough classification)，至於網路商店提昇交易品質的決策方案，為可控制條件變數，需符合原樣本空間中的決策規則，其與新決策之間的對應關係，即為提昇交易品質之作為的依據；本文亦將討論新決策對原系統的提昇擾動情況，以找尋新決策能力的邊界，並驗證導入約略集理論的優劣點。

關鍵字

電子交易品質 ； 約略集 ； 不可辨識關係表 ； 約略分類

並列摘要

For the e-commerce, providing the best quality and responding to customers' needs in time would be the key factor to increase the satisfaction and loyalty of customers. In this paper, the e-TransQual measurement developed by Bauer et al. (2006）would be applied by using the rough set theory to upgrade the transaction quality of e-commerce. The attribute dimensions and measuring quality elements of the e-TransQual information system would be assigned to be the attribute set and the universe set of the rough set respectively. The decision mapping process and the new decision-making methodology for the e-quality of on-line shopping would be carefully reviewed and deduced in this research.Specifically, the nature of the rough set theory is to clarify implicit data and ambiguity information, and finds out the appropriate decision for the information system. In this paper, the quality measuring factors would be the universe set, and 5 elements of functionality, enjoyment, process, reliability, and responsiveness would be the attribute set. Based on the questionnaire of the sampling space, the indiscernibility relation table could be constructed. By applying the Boolean algebra, the discernibility relation functions would be derived, and the core and the reduct attributes would be obtained. The process is called ”rough classification”. Therefore, the controlled new decision would be introduced to upgrade the e-TransQual information system by following original decision algorithms. The upgrading perturbation rating of the new decision would be derived, and the simulation result would be discussed to examine the validity of using the rough set theory.

並列關鍵字

Electronic transaction quality ； Rough set ； Indiscernibility relation table ； Rough classification