資訊管理學報

邱昭彰; 楊子青;
頁: 19-38
日期: 1998/01
摘要: 決策過程中,若能納入學習元素,並針對基本知識不足的決策者予以加強,將能增強決策者的分析能力及提昇決策支援系統之價值。唯有透過追蹤使用者與系統對話的過程來獲知使用者的目標·計劃等相關資訊,才能適時調整人機互動策略,以適應不同程度的使用者。因此本研究在超媒體環境中,提出了一個二階段的適性學習環境,以輔助財報分析之活動。在「學習階段」中,本研究的目標放在如何根據使用者的能力及進步情形來調整學習的內容,因此提出「複合節點:的設計概念,再結合案例庫推理與模糊邏輯所建構的「使用者模式推論:來決定適合使用者閱讀的內容深度,透過「節點內容處理」調整節點內容以組織成一適合使用者的超媒體環境。至於C測試階段」則加入了測試題目的出題·以及案例庫推理的技術,來共同判斷使用者是否已經充分學習,進而增進使用者之學習效果。藉由利用人工智慧相關技術本系統雛形已建構完成,期以提昇一適性學習與支援決策之環境。
關鍵字: 案例庫推理; 模糊邏輯; 財報分析; 過程追蹤; 智慧型教學系統;
Abstract: 工t's believed七hat七he provision of learning elements and decent instruction would improve a user's decision making capability during the decision一making processes. Only when a system is able to capture a user's problem一solving behavior during user一system interaction, can a system勿namically adapt to its users with different skill levels appropri-- ately. This research provides a two一stage adaptive multimedia-based learning environment to aid a user when conducting financial statement activities.In the 'Instruction" stage the system determines a user's initial skill level based on the user's profile. A compound nodes design concept is developed and integrated with case一based reasoning/fuzzy logic techniques to construct user models based upon which the appropriate learning contents are deter-mined. In the ',Testing" stage the system generates sets of financial statements scenarios
Keywords: Case-based reasoning; Fuzzy logic; Financial statement analysis; Process tracing; Intelligent tutoring system;;

瀏覽次數: 7543     下載次數: 43

引用     導入Endnote