林育志,黃正魁;
頁: 103-132
日期: 2022/04
摘要: 本研究利用項目為基礎的協同過濾想法，提出一種新穎的協作推薦模型。模型根據主觀查詢和客觀規則，透過關聯法則和相似度演算生成推薦結果。並於參與者使用協作推薦系統後，藉由用戶體驗問卷量測使用者對模型的感知有用性、信任度和滿意度。我們以台灣50成分股作為實驗標的來收集真實數據集。根據研究結果，新穎協作推薦模型（系統）呈現出更高的感知有用性、信任度和滿意度。
關鍵字: 協作推薦系統 ； 關聯法則 ； 相似度 ； 主觀視角 ； 客觀視角;

Novel Cooperative Recommendation Model from Subjective and Objective Perspectives

Abstract: This study utilizes the idea of item-based collaborative filtering to propose a novel cooperative recommendation model. The model adopts the technique of association rule mining and the similarity computation algorithm to generate recommendations from subjective inquiries and objective rules. In addition, a user-experience questionnaire is conducted to measure the perceived usefulness, trust, and satisfaction after participants use the cooperative recommendation system. The experiment adopts the shares from the Taiwan Top50 Exchange Tracker Fund (ETF50) as recommendation items to collect our real-life dataset. According to the result, the novel cooperative recommendation model (system) presents higher perceived usefulness, trust, and satisfaction.
Keywords: Cooperative recommendation system ； Association rule mining ； Similarity ； Subjective perspective ； Objective perspective;

瀏覽次數: 13668     下載次數: 9998

引用     導入Endnote