專家系統自1965年發展至今，其發展是與日遽增，在邁入二十一世紀這個新紀元，傳統的專家系統遭遇到不同以往的問題，不僅處理的問題複雜度提高之外，在建置系統的過程中需要更多專家提供其寶貴的意見，以期讓系統在處理問題的層面能更加寬廣及增進其彈性和效用。因此，多專家系統能夠解決傳統上單一專家先天上的限制。Gaines和Shaw於1989年在其論著中指出，利用一群專家的知識來發展專家系統其效益比單一專家來的更好。 然而，在多專家的專家系統中會產生案例選取的衝突，過去大都依賴人為的經驗法則判斷。如此一來，不僅在處理的時間成本上產生耗損，其選取案例的公信力亦容易令人產生存疑。此外在相關的研究上，絕大多數都未對於此一衝突提出另外一套較具公正性的解決辦法。 針對此一現象，本研究發現欲解決其中之案例衝突，可以藉由群體決策和多評準決策領域中尋求解決之辦法，透過本研究一連串的文獻蒐集與探討，得到Nemawashi決策模式可以加以導入應用，因此，本研究嘗試引用案例式推論（Case-Base Reasoning）、Nemawashi 決策模式，提出一個整合多專家的意見和解決其案例產生衝突的方法。 / Expert system has been in speeding development since 1965. With the advent of the 21st century, the traditional expert system is encountering problems different from the past. With the rising complexity of nowadays problems, it requires valuable opinions from more professionals in the construction of expert system. The multi-expert knowledge can not only broaden the scope in which the system handles problems, but also enhance the system’s flexibility and efficiency. Thus, multi-expert system outsmarts the conventional expert system which is restricted by the voice of a single expert. Gaines and Shaw in 1989 commented that the expert system was better quipped with a group of experts than with one single expert. Nevertheless, multi-expert system contains the problem of case conflict. To undermine the conflict, it is common for people to resort to experts’ judgments and their personal experiences. Consequently, the multi-expert system has the disadvantage of consuming time in the process of case selection. Moreover, the case selected out of this process may be unconvincing due to its overdependence on human decisions. As to the problem of case conflict, most of the researches related to multi-expert system do not propose other more objective solutions. Focus on the problem mentioned above, this study tends to solve case conflict through the use of Group Decision and Multiple Criteria Decision-Making (MCDM). After the collection and analysis of data, the study finds out that Nemawashi Decisions are effective in handling the problem of case conflict. Thus, this study attempts to apply Nemawashi Decisions in Case-Base Reasoning in order to combine opinions from different experts and to solve the case conflict in the multi-expert system.

臺北都市公有空間由於缺乏供給與需求配置的妥善評估，產生許多設施與建築發生閒置與低度利用的情況，造成環境的負擔以及資源的不平均分配。同時，臺北市持續面臨高齡化社會現象，銀髮福利設施是政府未來發展項目重點之一，現行多數研究先少探究銀髮福利設施於區域之設置標準。本研究針對公有閒置設施活化利用為銀髮福利設施，透過國內外文獻回顧，歸納「安全性」、「規模性」、「參與性」為銀髮福利設施主要活化原則，並建立分析模型。接著援用各級政府閒置設施資料庫，並結合社經資料，應用GIS系統進行閒置設施供給與銀髮福利設施需求的配置分析。透過分析模型與多評準決策分析法，篩選與評估不同地區與空間適合發展銀髮福利設施配置，藉此研擬相關活化策略。 實證結果顯示臺北市市中心位置對於銀髮福利設施之需求為高，相對其缺乏資源，以萬華區、文山區、松山區為嚴重。各公有閒置設施受到內在外在條件之影響，各閒置設施發展強度不同，結果顯示鄰里型服務範圍較適合發展安養護機構、社區型服務範圍較適合發展日間照顧中心、地區型服務範圍較適合發展安養護與日間照顧混合型設施。最後依據各別公有閒置設施之配置結果，建議未來公有閒置設施活化為銀髮福利設施之策略，配合設施內在外在條件，如人口、政策、文化、區位等等，使閒置設施達到最大活化效益。 / In Taipei City, there are government owned urban facilities and buildings which become idle facilities and with low utilization due to the lack of proper evaluation of supply and demand. It causes uneven distribution of resources to environment. Meanwhile, Taipei city faces the issue of aging society. The Government has to develop pension facilities for future planning. There are few existing studies discussed about the configuration standard of pension facilities. Therefore, this study explores the planning principles of pension facilities included safety, scale of facilities, participation, and the activating operations of public idle spaces by using the social-economic and idle spaces data of the Government. Also, this study simulates supply and demand data with GIS system. We analyze and evaluate the allocations which suitable for developing pension facilities in different regions and spatial conditions through Allocation Simulation, multiple criteria decision. Thus, it can be a reference for activating idle spaces for Taipei city. The results of supply and demand simulation show that the demand for pension facilities is higher in the center of Taipei City that lack of resources, including Wanhua District, Wenshan District and Songshan District. Each government owned urban facilities may be affected by outer and inner conditions which cause different development intensity. The results show that neighbor-type facilities are suitable for developing retirement centers, community-type facilities are suitable for developing day care centers, region-type facilities are suitable for developing mix-used centers. Finally, according to all allocation results, it is recommended strategies should be added with outer and inner conditions which including population, policy, culture, location and allocation, in order to maximize the usage efficiency of idle facilities.