小世界流行病學建模與公共衛生政策評估：利用社會分身點概念與區域資訊建構社會網路式流行病學電腦模擬

Abstract

本論文的主旨是探討電腦建模與模擬在社會科學領域裡的重要性、優缺點、運用方式、各項驗證標準與評估程序。本論文擬以電腦建模與模擬為核心重點，詳細介紹二項原創性的社科模擬研究。首先，第一項研究運用電腦建模與模擬來探討流行性傳染病的傳播動態。此研究提出一套新穎的社會分身點概念來表示長距離移動、日常定點活動與多個活動點等重要的社會現象。並結合細胞自動機，建構一個能夠描述日常生活接觸與人際互動模式的小世界社會網絡模擬模型。此研究模擬2003年初在台北、新加坡與多倫多（加拿大）三地的SARS疫情，模擬結果與實際的疫情資料具有高度的一致性，充分吻合當地疫情的發展趨勢與特色。另外，此研究也證明該模擬模型非常適合探討與流行性傳染病有密切相關的公衛政策與防疫策略等。其次，本論文在第二項研究中，針對近來非常熱門的小世界社會網絡模擬模型，提出一個傳播問題模型，並設計一個敏感度分析實驗，來探討數種區域差異對於模擬歷程與結果的影響，並藉此檢視各種區域差異的影響力與敏感度。最後，本論文期望透過上述二項研究來充分掌握電腦建模與模擬的特性與潛力，釐清它的操作程序與潛力，進而探討它的意義、限制與應用範疇。

The purpose of this dissertation is to explore the importance, advantages, applications, validation standards, and evaluation procedures of computational modeling and simulation in the social sciences. I will present two original social simulation studies in detail. In the first, computational modeling and simulation are used to investigate the transmission dynamics of epidemics and to evaluate the effectiveness of various public health policies and epidemic prevention strategies. A novel social mirror identity concept is proposed to represent social phenomena such as individual long-distance movement, daily visits to fixed locations, and multiple activity locations. The cellular automata concept is also utilized to construct a small-world social network model that represents human interactions and daily contacts. To test these concepts, I simulated the 2003 SARS outbreaks that occurred in Taipei, Singapore, and Toronto. The simulation results, which were highly consistent with actual epidemic data, corresponded with local outbreak trends and features. The simulation model was shown to be suitable for investigating public health policies and epidemic prevention strategies. In the second study I investigated the influence of local information on social simulations based on a small world model. I introduced a cellular automata-based variation with added shortcuts as a test platform for simulating the spread of an epidemic disease, then examined the influences of various local information factors on the results. It is my hope that these efforts will help future researchers determine appropriate simulation parameters, clarify operational procedures, and access meaning from simulations.