近年來,由於知識增加之速度激增,各類刊物/書籍發行之速度與數量亦隨之遽增;而讀者面對越來越多之書籍/刊物,如何從中取得其真正需要之資訊,並節省大量書籍/刊物過濾與篩選時間,實為讀者之一大需求。此外,各書籍零售業者亦希望增進讀者於門市內選購書籍之時效,並期望讀者可於圖書門市內獲得更多相關書籍之資訊,以提升讀者購買書籍之慾望,進而達成增進銷售業績之效果。因此,本研究乃以書籍之相關資料(如書籍基本資料、書籍銷售資料等)為基礎,建立一合理之書籍相關性推論模式,以取得相關書籍之資訊。取得相關書籍推論結果後,並以一推薦系統作為媒介,將符合讀者所需之資訊依其相關性高低呈現予讀者,供讀者作為選購書籍之參考,以達到促銷書籍之目的。此外,由於資訊技術(如網路技術與無線射頻技術)之快速發展,為提升該推薦系統使用之便利性與時效性,本研究乃結合無線射頻辨識系統(RFID)於此推薦系統內,期望藉由RFID系統無需與物體接觸即可感應商品識別標籤及可長距離感應識別標籤之特性,將其應用於協助讀者搜尋圖書門市內目標書籍與相關書籍所放置之架位,以減少讀者於門市內搜尋書籍所花費之時間。 本研究最終乃建立一套整合RFID技術之書籍推薦系統,並以一案例驗證此模式與技術之正確性與可行性。整體觀之,藉由本研究所發展之推論模式與技術,除可利用電腦所具之運算能力進行書籍相關性運算而協助讀者取得相關書籍之推薦外,亦採用先進RFID技術協助讀者進行書籍架位之搜尋;如此一來,將可達到便利讀者與提升書籍零售業銷售業績之雙贏局面。
Recently, the releasing rate and amount of all sorts of publications/book issues have been significantly increased due to the drastic development of domain knowledge. Since readers have to face more and more publications/book issues, to efficiently acquire the useful materials from a large quantity of publications/book issues has become a critical issue. In addition, the book retailers expect to improve the efficiency for readers to select and purchase books at the stores and seek to obtain good business performance. Thus, based on book-related attributes (e.g., the basic descriptions and sales information), this research develops an algorithm and a recommendation system to explore the correlation of publications in order to provide readers useful recommendation for purchasing books. Furthermore, concerning the automatic identification capability of RFID, this research integrates the RFID technique in the proposed recommendation system to improve the performance of book searching and recommendation. In addition to the proposed algorithm and recommendation system, a demonstration scenario is also provided to analyze the feasibility and performance of the proposed approach. It is the goal of this research to develop an applicable model to enhance the sales of book retailers.