| DC 欄位 |
值 |
語言 |
| DC.contributor | 企業管理學系 | zh_TW |
| DC.creator | 陳佩琦 | zh_TW |
| DC.creator | Pei-chi Chen | en_US |
| dc.date.accessioned | 2014-7-29T07:39:07Z | |
| dc.date.available | 2014-7-29T07:39:07Z | |
| dc.date.issued | 2014 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=101421021 | |
| dc.contributor.department | 企業管理學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 隨著經濟發展，消費者的需求越來越多元，不同消費族群對於商品或服務的需求及對其願付價格不盡相同。在這樣競爭激烈的環境下，企業已不容易將目標市場的所有顧客都攬進自己的市場內。此時，找出對企業有價值的重要顧客成為企業成功獲利的首要步驟。找到重要顧客後，若是能夠發現對重要顧客而言具有價值的商品，即可投其所好，透過滿足關鍵顧客的需求而達到長期盈利的目的。
本研究使用台灣某中型零售商某年度之年度交易記錄，利用RFM價值分析法將有顧客編號之交易資料者分別賦予R、F、M三個指標的分數後，利用K-means分群法將顧客價值由高至低分成五星等至一星等。接著本研究使用改良版Apriori演算法透過最小支持度以及偏態係數兩項條件之設立進行關鍵商品之探勘。將候選項目集合利用最小支持度和偏態係數做驗證，將同時符合兩條件者列為我們所要找尋的關鍵商品；只通過最小支持度而未通過偏態係數者列為不完整項目集合，並將不完整項目集合合併為下一階段之候選項目集合以進行下一階段最小支持度及偏態係數的探勘，以此類推，直到無法在下一階段中找出任何高頻項目集合為止。過程中，不滿足最小支持度之驗証者直接刪除，並不再進行偏態係數之驗證。本研究希望能透過這樣的探勘過程找出對企業貢獻度高的重要顧客經常購買的關鍵商品(組合)，進而以此關鍵商品(組合)作為評斷未來新顧客是否為潛在重要顧客之基準。
本研究最後找出了31個單一品項關鍵商品以及60組兩個品項的關鍵商品組合，且在資料驗證部分中得到高達69%的驗證比率。
| zh_TW |
| dc.description.abstract | As economy developing, the demand of consumers getting more and more diverse. Each consumer cluster has its own demand of the product or service and the price they are willing to take. Under such a competitive environment, it is not easy that enterprises occupy the whole target market any more. Figure out who are those important customers, however, be-come the first step to follow for making profit. If it is possible to find out what means the valuable products for those important customers, catering to their pleasure, enterprises can achieve the purpose of making long-tern profit through meeting those key customers’ need.
The study applies one of the annual transaction record of a medium-sized retailer in Taiwan. Given the transaction data tagged by customer numbers three indicators scores-R,F,M by us-ing RFM analysis method. Then the customer value will be divided low into five degrees by K-means cluster method. The n the study starts mining the key product though the setting of two conditions- minimum support and coefficient of skewness by the modified Apriori al-gorithm.
To verify the candidate itemsets by minimum support and coefficient of skewness and clas-sify the ones which meet both of the two conditions as the key product. The other ones that only meet the minimum support but not the coefficient of skewness will be classified as in-complete itemsets, combined to be the candidate itemsets for the next stage mining and so on until that finding out any other frequent itemsets is unable to be done. The others that don′ even meet the minimum support will be deleted right away. This study wants to find out the key product(s) that the VIP customers buy frequently through this kind of mining process. And then see the key product(s) as a basis of determining if the new customer is a potential VIP or not.
The study identified 31 single key product as well as 60 pairs of key products eventually and got up to 69% of the validation ratio
| en_US |
| DC.subject | 資料採礦 | zh_TW |
| DC.subject | data mining | en_US |
| DC.subject | apriori | en_US |
| DC.title | 尋找關鍵商品組合之研究 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Study on finding key itemsets | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |