- 學位論文
數據挖掘技術在圖書推薦採購的應用
Application of Data Mining technology to recommended book purchasing
摘要
數據挖掘(簡稱DM)技術已經充分應用於各領域，而在圖書館應用方面也有不少的文獻探討過。但就圖書推薦來說，之前大多數的論文都在強調利用不同的方法，推薦圖書給讀者，希望藉此來提供更好的服務或是提升館藏的使用率。 本研究主要是以國立中正大學圖書館自動化系統3年的紀錄資料進行數據挖掘，經過事先的資料預整理，擷取出有效的欄位，最後整合成單一的表格資料後，匯入SPSS軟體進行分析，產生可供迅速決策的決策樹。 雖然與之前大多數的論文一樣，是以流通資料為數據挖掘的來源，但卻以不同的角度切入。希望經由SPSS統計軟體的DM技術，產生出決策樹，針對讀者推薦購買的新書，給予迅速的決策，回應其是否購買。如此一來不僅能夠減少館員每天的查核工作，讓其專注於更重要的服務與工作，也能降低讀者的抱怨。更能減少館員主觀意識所產生的影響，進而提升圖書採購的品質、讓圖書採購經費的有效配置，進而提供更好的服務與提升館藏的使用率。
並列摘要
Data mining (DM) technology has been widely applied to various fields, and there were many applications for library in many literatures. As for book recommendation, most literatures emphasized that using different methods recommended books for users in order to provide better services or to enhance the utilization of collection. This research mainly uses the 3-year record of data in the library automation system of Chung Cheng University, and gains a data table after data pre-processing and extracting the effective fields, then finally imports the data table into SPSS software and rapidly produces the decision tree. Like the many previous literatures, the research uses the circulation data for data mining, but considers from a different point of view. It uses the DM technology in SPSS to produce the decision tree, and rapidly gives users the answer whether purchases the books those they recommended or not. Not only decreases the amount of everyday checking work of the librarians, let them can focus on more important service and work, but also reduces complaints from readers; and further, reduces the impact of subjective awareness of the librarians to increase the quality of purchasing books and allocation of purchasing fund, and finally to provide better service and enhance the utilization of collection.