Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/92689
標題: 應用資料探勘方法分析顧客價值 - 以台灣某化學產業為例
Application of Data Mining on Enterprise Customer Vaule – A Study of Some Chemical Industry in Taiwan
作者: Jiunn-Jer Chang
張俊哲
關鍵字: RFM model
K-Means model
UV curing
Customer Value
RFM模型
K-Means模型
紫外光固化
顧客價值
引用: 參考文獻 中文部份: 工研院IEK(2014)。光固化市場調查。 尹相志(2006)。SQLServer2005資料採礦聖經。台北:學貫出版社。 林祥生、劉益豪(2008)。應用資料採礦探討國際線航空旅客之線上購票行為。運輸計劃季刊,37(2),197-235。 金養智(2009)。光固化材料性能及應用手冊。化學工業出版社。 郝沛毅、李御璽、黃嘉彥編譯(2014)。資料探勘(3/e)。ELSEVIERTAIWANLLC高立圖書合作出版。 陳民祐(2012)。應用資料探勘探討線上訂票乘客行為以台鐵西幹線為例。國立中興大學行銷系碩士論文。 陳亭羽、賀千盈(2005)。應用決策樹探討適用於電子行銷市場之區隔基礎。管理科學研究季刊,2006(p1~25)。 黃元直(2009)。RFM模型區隔消費者購買行為的區別能力研究-保險銷售資料實例驗證。私立德明技術學院論文。 黃壬暉(2013)。回應式鐵路車票預售策略之探討。國立中興大學行銷系碩士論文。 蘇冠豪(2006)。應用資料探勘技術於醫療行銷之研究—以醫學美容為例,國立雲林科技大學資訊管理系碩士論文。   英文部份: BabakSohrabi,AmirKhanlari(2007).CustomerLifetimeVaule(CLV) MeasurementBasedonRFMModel.IranianAccountin&Auditing Review,Spring2007,Vol.14No.47,7-20. Blattberg,R.C.,Briesch,R.,&Fox,E.J.,(1995)'HowPromotionsWork',MarketingScience,(14),122-132. CharlotteMason(2003),'Recency,FrequencyandMonetary(RFM)Analysis'http://cmason.myweb.uga.edu/Course_Roadmap/_RFM_Analysis/Note_-_RFM_Analysis.pdf' ClaudioMarcus(1998).Apracticalyetmeaningfulapproachto customerSegmentation.JOURNALOFCONSUMERMARKETING,VOL15 NO.51998,494-504. D.Hand,H.Mannila,P.Smyth.(2001)PrinciplesofDataMining.MITPress,Cambridge,MA. Drucker,Peter(1954)'ThePracticeofManagement'. Han.W.&M.Kamber(2007).DataMiningConceptsandTechniques.70-77. Hughes(1994).StrategicDatabaseMarketing.Chicago,ProbusPublishing,JiaweiHan,MichelineKamber(2007).'DataMiningConceptsandTechniques',70-77. Jo-TingWei,Shih-YenLinandHsin-HungWu(2010).AreviewoftheApplicationofRFMmodel.AfricanJournalofBusiness ManagementVol.4(19)4199-4206. JohnA.McCarty,ManojHastak(2006).Segmentationapproachesin data-mining:AcomparisonofRFM,CHAID,andlogisticregression.JOURNALOFBusinessResearch60(2007)656-662. MacQueen,J.B.(1967).Somemethodsforclassificationandanalysisofmultivariateobservations.UniversityofCaliforniaPress,1,281-297 Mason,CharlotteH.(2003),'TuscanLifestyles:AssessingCustomerLifetime Value,'JournalofInteractiveMarketing,17(Autumn),54–60. W.FrawleyandG.Piatetsky-ShapiroandC.Matheus(Fall1992).'KnowledgeDiscoveryinDatabases:AnOverview'.AIMagazine:213-228 Zeithaml,V.A.(1988).Consumerperceptionsofprice,quality,andvalue:Ameans-endmodelandsynthesisofevidence.JournalofMarketing,52(July),2-22
摘要: 摘要 在這個資訊充足的時代裡,如何利用現有資料,去做分析,以取得一個整合性資料,進而分析得到一個關連性的資料,而以此資料來做未來之預測,這是本研究的目的。因現代的產品生命週期相當的短,故在投入資源時,如何在有限的資源下,投入對的產業或客戶,相對來說也就越來越重要了,因此如何在利用過去的資料,發展出一套好的行銷模式,來尋找最有利基的客戶群,所以依照過去的交易資料,做客戶分類後,再分析其客戶的特性,研擬不同之行銷策略。 此研究利用RFM(P)和K-Means分析法作一比較後,擇一較適用的方法,作為分類的依據。利用此分類法,將客戶分為數群後,再依此分類出的客戶群,作為後續針對不同組群,做不同之行銷策略。故本篇研究在於如何利用資料探勘的方法,找尋最有顧客價值的企業用戶。 研究結果利用RFM(P)分析法較為適用,先利用顧客價值將其分為4組,將其4組分為2群(高顧客價值與低顧客價值);於此2群中,配合行業別將其分為三組,此三組內各含有不同行業別,其中高顧客價值的有1,4,6三個行業,其次是2,3,5,8,9五個行業,最低階顧客價值之行業則有7,10兩個行業。以此以上三組分類,再擬訂不同的行銷策略及投入資源多寡的依據。
Abstract In this age of information explosion, it is important to make the best use of accessible information by analyzing and integrating it into continual information for the use of future prediction. It is the study's aim to attest the best approach for doing it. Considering the short life cycle of today's products, it is comparatively essential to single out the right investees with limited resources, whether the customers or the industries. Therefore, a company must be able to identify the most potential clients by searching through the historical trading records before the marketing strategies are made in accordance to the characteristics of each of the groups of clients. This study compared two approaches, RFM (P) and K-Means, in order to decide which the most useful classification criteria was. The next step was to divide the clients into several groups and develop differing marketing strategies accordingly. In other words, it was the objective of the study to find out the best way for a company to identify the enterprise users of highest customer value by exploring accessible information. This study attested the approach of RFM (P) as most applicable. The clients were divided into four groups by customer values, grouped two by two (as groups of high/low customer value), and then categorized according to the trades as three aggregations. There were various trades among the three aggregations. As the results showed, there were trades 1, 4, and 6 in the first aggregation (of highest customer value), trades 2, 3, 5, 8, and 9 in the second, and trades 7 and 10 in the third one (lowest customer value). The last step was then to draw up differing marketing strategies and to determine the appropriate amount of money to be invested correspondingly.
URI: http://hdl.handle.net/11455/92689
其他識別: U0005-1007201512445000
文章公開時間: 2018-07-21
Appears in Collections:高階經理人碩士在職專班

文件中的檔案:

取得全文請前往華藝線上圖書館



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.