本計畫以動態社會網路分析及多關聯多層級網路模型為兩大研究核心議題，並以政府及政黨高層職位之異動為資料領域，探討政治權力變化之指標、現象與趨勢。在第一階段的研究中，我們希望透過政府官員職位異動資料，分析各部會的專業化程度，利用動態社會網路分析的方法，將政府組織中個體的異動事件加以凝聚成群體的變化事件，再依據群體的事件變化趨勢，以專業化程度、部門性質導向、升遷特徵及趨勢、以及核心人物之專業群組，來檢視及探討從觀察中獲得的重要資訊。計畫的第二階段則延伸至以多關聯及多層級之動態模型觀點，探討政治權力提升模式及影響的關鍵因素。我們將列舉現實世界下各種層級可能之政治權力變化相關行為模式，並且以多關聯(multi-relational)的概念給予這些行為模式一合理且具備鑑別力的權重數值，從中觀察篩選取得關鍵事件(critical event)，並以多層級(multi-layered)觀點來設計對應之量化指標模型，藉以關鍵指標模型的量化結果來重新塑造並盡可能貼切於真實之政治環境互動。 本計畫亦將參考政治資訊學(Political Informatics)、生態政治學(Political Ecology)、社會資產(Social Capital)的模型以及近年趨勢，定義出政治權力變化之網路節點與結構指標，試圖以一資料整合(data integration)的觀點來盡可能捕捉現實世界中相關的事件、資料或資訊，並藉以前述模型的雛型方法運作(prototyping)結果，交互驗證並藉以做為修改的依據。 The research project proposes to address the advanced topics of dynamic social network analysis and multi-relational, multi-layered network model, using executive positions of government agencies and political parties as data domains. The overall objective is to develop appropriate measures of political power transition and provide a platform for political power observation. In the first stage of the proposed research, we intend to analyze promotion behavior and power group in government agencies. Personnel changes are modeled as events in a dynamic social network. Individual events are aggregated and transformed into triggers of community evolution. This observation of community evolution provides identification and important insights into characteristics of political power group. The second stage of the project will extend the investigation to a more complex model of multi-relational and multi-layered network. The flexibility of different weights in edges and different types of nodes leads to a refined modeling of political power operational behaviors in a realistic context. As a result, distinctive measures and critical events can be better derived so as to allow a fine-grained and accurate observation. In the process of conducting the research project, we will also draw upon recent advances in the fields of political informatics, political ecology, and social capital in developing dynamic social network model for political power evolution. This involves a mapping of political power operations to definitions of network nodes and edges. Measures of both nodal and structural characteristics are also utilized and revised to reflect actual phenomenon in political power evolution. The dynamic network model of political power evolution will serve as a platform for data integration that captures relevant data and events. Finally, we will develop a prototype system and perform iterative experiments and improvement.