以鐵路為譬喻之資訊演化可視化研究

Abstract

本篇論文提出一新穎的可視化研究：鐵路視覺化（Railway Visualization），以表達隱含時間關係的資訊演變。本研究可有效地表現個體間的內部狀態演化；與一段時間內不同個體間的關係變化，如分歧、交錯、聚合、共行、平行。藉由使用生活中的地鐵圖為譬喻，我們可利用人類對於地鐵圖的心智模型，並兼顧schematic map與route map的優點，以幫助使用者有效且輕鬆地探索資訊的演化關係。同時，本研究也提出地鐵繪製的美學原則，並利用兩步驟多重規則的最佳化演算法，為抽象資料自動產生地鐵圖風格的視覺化圖形。除了展示論文、電影、網路相本、網路文件等四種資料的應用結果，我們也和傳統圖形視覺化表示相比，討論二者的優勢與適用範疇；並總整過去的自動生成地鐵圖演算法的美學原則，且提出本研究的泛用與應用領域。

In this thesis, a novel graph representation, Railway Visualization, is proposed to reveal the information evolution with temporal relationship over time. Through Railway Visualization, it not only reveals the changes of an entity’s intra-state over time, but also effectively discloses the intrinsic characteristics of the complicated relationship evolution, such as branching, crossing, converging, merging and parallelism. Besides, this work can leverage people’s existing knowledge and mental models to quickly interpret the graph and gain the insight by adopting the railway route map in daily life as the metaphor to visualize the information evolution. Moreover, by observing the railway route maps, the drawing aesthetic rules are concluded and formulated as the energy terms. Then, this drawing problem is solved by the two-steps optimization. Furthermore, the interactive level-of-detail mechanism is also provided to explore data according to users’ interests. As a result, several types of data, including papers network, movies, photo albums and news articles are visualized to prove the virtues of the Railway Visualization. Finally, the generalization, strength and limitation of this work are compared with the traditional graph visualization.