以合作賽局角度看待慈善機構與個案之間的配適問題

Mechanism Application to Match Charities Organizations and Donees Based on Cooperative Game Approaches

目前慈善機構與個案端間的資源分配研究較少，在實務上也常看到照顧個案資源分配不均，有些照顧個案接收到同質性的福利物資過多，或是更需關懷個案卻無資源協助之現象，本研究利用合作賽局機制「延遲接受演算法」設計一個配適平台，作為慈善機構與照顧個案之間的媒合，慈善機構間可因資訊互通及透明化，使資源更有效利用，進而合作擬訂關懷計畫。本研究模擬不同情境，結果顯示「延遲接受演算法」之整體成效優於現行的配對機制與「波士頓配對機制」。本研究機制亦可作為社會工作人員決策之參考，以達到慈善機構之價值。
Few studies focus on the matching problem between charities organizations and donees. In the real world, there is an uneven distribution phenomenon among donees. Somet donee cases receive too much charity resources, while other cases with higher priority may receive no charities. This study applied a ''deferred acceptance algorithm'' to design a matching platform for charity organizations and donees. This platform makes the charity distribution process both logical and transparent| therefore, the charity organizations will distribute their resources more efficiently. In the three scenarios designed in this study, all the results support the conclusion that the deferred acceptance algorithm has better performance than the current matching mechanism and Boston matching mechanism. This platform can also serve as a reference for social workers in order to achieve the desired value of charities.