3D Information Visualization Based on Spatial Deployment Strategies

[[abstract]]現今，網路上迅速累積大量的資料，這些龐大的資訊時常會造成使用者在搜尋與閱讀上的混淆，如何透過資料視覺化系統，幫助使用者發掘其中與搜尋關鍵字真正相關的資料，這個議題已經成為十分重要的研究課題。因此，在本研究中我們開發一個新的系統，其目的在於透過資料視覺化與多層解析的3D介面增強使用者感知。 我們的系統使用資料探勘技術，它可以透過網路爬蟲蒐集網路上相關的評論資料，並且解析網頁上相關的關鍵字再構成階層化資料結構。考量到使用者在同一時間內能夠辨識目標的數量限制，系統根據Octree空間分配概念建構出階層化資料結構，在資料結構中同一階層的節點各有一個相對的3D座標位置，並且透過force-driven方法去計算與佈署各節點在介面中的位置，處理後的成果會以互動式的圖形介面呈現給使用者，這樣的方法不但能夠保持介面的易讀性，也能夠並提供使用者與過往3D經驗一致的視覺提示，實驗結果顯示，我們系統能夠有效提升使用者的感知，並且提供一個潛在的機會發現。[[abstract]]Nowadays, information on the internet has accumulated so rapidly and explosively. The tremendous amount of information often causes users confusion while searching and browsing. How to help users find really relevant data in an information visualization system has become an important research issue. Therefore, in this research, we develop a novel system which aims to intuitively enhance users’ perception by visualizing data and its correlations through a multi-resolutional 3D interface. By utilizing the data mining technologies, our system analyzes related keywords, which are extracted from the related posted comments on the Internet by a Web crawler, and clusters them together into a hierarchical structure. Considering the limitation of the number of targets that a user can recognize simultaneously, our system constructs the data hierarchy according to the Octree Space Partitioning (OSP) method. And the relative 3D position of each node on the same level is calculated and deployed by using a force-driven method. The result is then interactively represented through a graphical user interface. This approach does not only keep the layout of screen easy-to-view, but also provide users useful visual clues which are consistent with their 3D stereoscopic experience. The experiment results show that our system can successfully enhance users’ perception, and, moreover, provide the potential of a chance discovery.[[note]]碩