- 學位論文
應用二維度縮減案例式推論系統於休閒阻礙之分類
A case-based reasoning system with the two-dimensional reduction technique for leisure constraints classification
摘要
一、 研究動機及目的：本研究旨在發展一套植基於案例庫的休閒阻礙推論系統，協助休閒業者由數個人口統計變數推論潛在遊客可能遭遇的休閒阻礙型態，以利於休閒業者進行休閒阻礙協商及策略性行銷。 二、 研究方法及設計：本研究以月眉探索世界的潛在遊客為研究對象，經問卷調查法蒐集案例(樣本)後先將案例分群；再以案例式推論技術結合基因演算法，選擇具代表性的案例特徵(人口統計變數)及案例，完成案例庫休閒阻礙推論系統之建置。 三、 研究結果：蒐集之案例(樣本)可分為無阻礙、個人內在阻礙、高度阻礙及外部阻礙四群。而系統經基因演算法優化後，推論所需之案例特徵(人口統計變數)由原有六項簡化為三項。研究結果證實藉由三項人口統計變數本系統可有效推論出潛在遊客所屬之群體。 四、 研究貢獻：休閒業者對於不同潛在遊客可能遭遇的休閒阻礙型態通常不易掌握，而本研究首先應用人工智慧技術，可輔助業者掌握不同遊客可能遭遇的阻礙因素，此對於業者進行休閒阻礙協商及策略性行銷具有實質貢獻。
並列摘要
Purpose－This study aims to develop a case-based reasoning (CBR) system for leisure constraints. This system could infer leisure constraints the visitors perceive according to their demographic variables, and thus could offer marketing strategies of leisure constraints negotiation. Design/methodology/approach－The cases for building the CBR system were collected based on the survey of people who visited the Yamay Discovery World. Each case consists of features(demographic variables) and perceived constraints. Then, a cluster analysis was made to classify the cases into mutually different sets using perceived constraints. Finally, the CBR system was built up with the application of genetic algorithms(GAs) to feature and case reductions. Findings－The cases are divided mutually into four sets, namely no constraints(NC), intrapersonal constraints(IC), strong constraints(SC) and external constraints(EC). According to reduced features (demographic variables) the system can efficiently infer the set a new case (visitor) belongs. Originality/value－This study pioneers on applications of artificial intelligence(AI) techniques to inferring perceived constraints of visitors, which is usually not easy to catch on.