- 學位論文
視覺化情境式多維度瀏覽
Visualization Support to Contextual Multi-dimensional Browse
摘要
利用視覺化方法(visualization techniques)來呈現多維度資訊(multi-dimensional information)能幫助使用者更有效地瀏覽，這些方法包含了sliding rods、star coordinates和parallel bargrams都能清楚地呈現每個維度以及所有的項目。然而，這些方法也還是有些缺點，他們都主要著重在每個項目上，而沒有呈現出其他像是維度間關係的資訊，並且，這樣的呈現會受限於資料量大小及維度的數目而無法呈現大量的資訊。 因此為了解決以上的問題，本研究提出了一個hyperbolic-like點線圖的設計，其中每個節點代表一個維度。為了能選擇最好的呈現方式，我們整理了學者的研究關於不同圖形屬性(graphical attribute)傳遞不同型態資料的效果，之後我們對受試者做實驗，觀察他們對於不同圖形屬性有何不同的看法。經過一系列的實驗，我們發現利用節點顏色(node color hue)來呈現不同類別的維度、線段長短(edge length)呈現維度不同的重要性、節點顏色深淺(node color saturation)來呈現資料量大小為最有效。此外，我們也證明了圖形的呈現更能表現多維度資訊相對於以文字為主的介面。
並列摘要
To use visualization techniques to display multi-dimensional information helps users browse effectively. Among of them such as sliding rods, star coordinate, and parallel bargrams, present each dimension with all items clearly. However, these still can be improved. Most of them focus on an individual item without the relations between dimensions. Besides, the display is limited by the size of the data sets and the number of the dimensions. In this paper we propose a hyperbolic-like nodes and edges diagram, where a node represents a dimension. In order to choose the best display, we arrange previous studies of the effectiveness of the graphical attributes for conveying different types of information. Afterwards, we conduct a within-subject empirical study of the effectiveness of our conclusion. Through experiments, we find that node color hue, edge length, and node color saturation are effective to encode dimension classification, dimension importance, and data volume effectively. In addition, we prove that the graphical display is much better than the text-based one to present multi-dimensional information.
參考文獻
延伸閱讀
- 吳孟松（2010）。深度圖為基礎的立體視覺多視角影像生成〔碩士論文，國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU.2010.10016
- 陳俊源（2012）。巨觀且聚焦視覺通道實體資料繪製〔碩士論文，國立臺北科技大學〕。華藝線上圖書館。https://doi.org/10.6841/NTUT.2012.00657
- Chen, Y. J. (2011). 整合及呈現社群網絡平台上大量社群資訊之新穎視覺化技術 [master's thesis, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU.2011.00264
- Yeh, C. L., Lu, L. C., & Wu, S. H. (2010). Design and Implementation of Declarative Agent-Based Front-End System for ebXML. 電子商務學報, 12(1), 103-131. https://doi.org/10.6188/JEB.2010.12(1).02
- 來毓庭（2021）。Multi-Label Deep Visual-Semantic Embedding with Visual Transformer〔碩士論文，國立臺灣師範大學〕。華藝線上圖書館。https://doi.org/10.6345/NTNU202101778