頁籤選單縮合
| 題 名 | 利用分類分析發掘產品最適性之項目組合=Using Classification Analysis to Find the Most Adaptive Intemsets of Products |
|---|---|
| 作 者 | 陳垂呈; | 書刊名 | 資訊科學應用期刊 |
| 卷 期 | 3:2 2007.12[民96.12] |
| 頁 次 | 頁79-100 |
| 分類號 | 312.1 |
| 關鍵詞 | 資料探勘; 分類分析; 項目組合; Data mining; Classification analysis; Itemsets; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 本論文以消費者之交易資料為探勘的資料來源,每一筆交易資料包含消費者曾經購買的產品項目與其次序性,並以某一產品為探勘的目標,若交易資料包含此產品,則設定交易資料與此產品之間有關聯性,否則設定無關聯性。我們利用分類分析分別從以下兩方面發掘此產品最適性的項目組合:一是只考量產品是否出現在交易資料中,視其他產品項目為分類的影響屬性,將交易資料進行分類計算以建構出一決策樹。從決策樹之路徑中所顯示的關聯性,可做為發掘此產品最適性之項目組合的依據。二是增加考量購買產品的次序性,對包含此產品的交易資料,將其在購買此產品之前及之後的產品項目依序作分解,視分解後的項目組合為分類的影響屬性,將交易資料進行分類計算以建構出一決策樹。從決策樹之路徑中所顯示的關聯性,可發掘具有購買次序之此產品最適性的項目組合。我們依據所提出的探勘方法,設計與建置一個探勘系統,以發掘產品最適性的項目組合型樣。此探勘結果,對企業擬訂產品的行銷服務策略,可以提供非常有用的參考資訊。 |
| 英文摘要 | In this paper, we use consumers’ transaction data as the source data of mining. Each transaction data contains a consumer ever bought product items with sequences. We let one product as the target of mining, and assign the association between the transaction data and the product to be “yes” if the transaction data contains the product. Otherwise, it is “no”. The classification analysis is used to find the most adaptive itemsets of the product from two aspects: One is to consider whether items are contained in transaction data. We treat other items as attributes for classification, and classify the transaction data to construct a decision tree. We can find the association to be “yes” between some attributes and the product according to the paths of the decision tree. It is the basis to find the most adaptive itemsets of the product. The other is to consider extra the sequences of items in transaction data. If a transaction data contains the product, we have to divide sequentially the items before the product into itemsets, and the items after the product into itemsets. We treat these itemsets as attributes for classification, and construct a decision tree after classifying the transaction data. According to the paths of the decision tree, we can find the association to be “yes” between some attributes and the product. It is the basis to find the most adaptive itemsets of the product for considering the sequences of buying. The results of the mining can provide very useful information to plan the strategy of marketing services of products for the business. |
本系統中英文摘要資訊取自各篇刊載內容。