Fractal Approaches for Visualizing Huge Hierarchies Hideki Koike, Hirotaka Yoshihara Department of Communications and Systems University of Electro-Communications.

Slides:



Advertisements
Similar presentations
1 Greedy Forwarding in Dynamic Scale-Free Networks Embedded in Hyperbolic Metric Spaces Dmitri Krioukov CAIDA/UCSD Joint work with F. Papadopoulos, M.
Advertisements

Generated Waypoint Efficiency: The efficiency considered here is defined as follows: As can be seen from the graph, for the obstruction radius values (200,
Image Indexing and Retrieval using Moment Invariants Imran Ahmad School of Computer Science University of Windsor – Canada.
Self Organizing Maps. This presentation is based on: SOM’s are invented by Teuvo Kohonen. They represent multidimensional.
KD TREES CS16: Introduction to Data Structures & Algorithms Tuesday, April 7,
CS2420: Lecture 19 Vladimir Kulyukin Computer Science Department Utah State University.
A New Force-Directed Graph Drawing Method Based on Edge- Edge Repulsion Chun-Cheng Lin and Hsu-Chen Yen Department of Electrical Engineering, National.
CONE TREES: ANIMATED 3D VISUALIZATIONS OF HIRARCHICAL INFORMATION George G. Robertson, Jock D. Mackinlay, and Stuart K. Card Xerox Palo Alto Research Center.
1 Balloon Views of Source Code and Their Multiscalable Font Modes Chun-Cheng Lin and Hsu-Chun Yen Dept. of Electrical Engineering, National Taiwan University.
Branch and Bound Similar to backtracking in generating a search tree and looking for one or more solutions Different in that the “objective” is constrained.
LensBar – Visualization for Browsing and Filtering Large Lists of Data Toshiyuki Masui Proceedings. IEEE Symposium on Information Visualization, 1998 元智資工所.
Adaptive Web Caching: Towards a New Caching Architecture Authors and Institutions: Scott Michel, Khoi Nguyen, Adam Rosenstein and Lixia Zhang UCLA Computer.
Cone Trees and Collapsible Cylindrical Trees
Towards a Better Understanding of Web Resources and Server Responses for Improved Caching Craig E. Wills and Mikhail Mikhailov Computer Science Department.
DEAPS By: Michael Gresenz Austin Forrest. Deaps A deap is a double-ended heap that supports the double-ended priority operations of insert, delete-min,
Strategies for Implementing Dynamic Load Sharing.
Copyright © Cengage Learning. All rights reserved. 5 Integrals.
Extending TreeJuxtaposer Nicholas Chen Maryam Farboodi May 9, 2006.
9.1 Properties of Similar Figures Learning Objective: To use ratios and proportions to find measures of similar figures and use scale models to find dimensions.
Tree Structures (Hierarchical Information) cs5764: Information Visualization Chris North.
Parametric Modeling.
5-5 Similar Figures Warm Up Problem of the Day Lesson Presentation
H3: Laying Out Large Directed Graphs in 3D Hyperbolic Space Tamara Munzner Stanford University 元智資工所 系統實驗室 楊錫謦 1999/11/3.
FEATURE EXTRACTION FOR JAVA CHARACTER RECOGNITION Rudy Adipranata, Liliana, Meiliana Indrawijaya, Gregorius Satia Budhi Informatics Department, Petra Christian.
Similar Figures 4-3 Problem of the Day A rectangle that is 10 in. wide and 8 in. long is the same shape as one that is 8 in. wide and x in. long. What.
Using Graph Parsing for Automatic Graph Drawing Carolyn. McCreary, Richard O. Chapman, and F.-S. Shieh IEEE transactions on systems, man, and cybernetics-part.
Bellwork – 1/6/15. Unit 6: Section 6.1 Ratios, Proportions, and the Geometric Mean (Starts on Page 356)
An Introduction to the Resource Description Framework Eric Miller Online Computer Library Center, Inc. Office of Research Dublin, Ohio 元智資工所 系統實驗室 楊錫謦.
EQ: How do you use the properties of exponents and logarithms to solve equations?
GEOMETRIC SOLIDS 1 Similar Solids. SIMILAR SOLIDS 2 Definition: Two solids of the same type with equal ratios of corresponding linear measures (such as.
Information Visualization –III Treemaps and Fisheye Views.
IAT 814 Trees Chapter 3.2 of Spence ______________________________________________________________________________________ SCHOOL OF INTERACTIVE ARTS +
Interacting with Huge Hierarchies: Beyond Cone Trees Jeromy Carriere, Rick Kazman Computer Graphics Lab, Department of Computer Science University of Waterloo,
Guidance on expressing the Dublin Core within the Resource Description Framework(RDF) Eric Miller, Paul Miller, Dan Brickley Dublin Core Metadata Initiative.
智慧型系統實驗室 iLab 南台資訊工程 1 Evaluation for the Test Quality of Dynamic Question Generation by Particle Swarm Optimization for Adaptive Testing Department of.
1 Fundamentals of Robotics Linking perception to action 2. Motion of Rigid Bodies 南台科技大學電機工程系謝銘原.
Web Architecture: Extensible Language Tim Berners-Lee, Dan Connolly World Wide Web Consortium 元智資工所 系統實驗室 楊錫謦 1999/9/15.
Bell Work 1/22/13 1) Simplify the following ratios: a)b)c) 2) Solve the following proportions: a)b) 3) A map in a book has a scale of 1 in = 112 miles,
Stylization and Abstraction of Photographs Doug Decarlo and Anthony Santella.
Solve the following proportions. a = 9 b = 7 c = 6 d = 6.
Isometric Sketches (Material developed by Project Lead the Way)
© 2008 Autodesk Isometric Pictorials. © 2008 Autodesk Isometric means equal measure. Three adjacent faces on a cube will share a single point. The edges.
Low-Dimensional Chaotic Signal Characterization Using Approximate Entropy Soundararajan Ezekiel Matthew Lang Computer Science Department Indiana University.
FRACTAL DIMENSION. DIMENSION Point 0 Line 1 Plane 2 Space 3.
CHI 2003 – Visualization & Navigation1 Efficient User Interest Estimation in Fisheye Views Jeffrey Heer and Stuart K. Card 1 Palo Alto Research.
Innovative UI Ideas Marti Hearst SIMS 213, UI Design & Development April 20, 1999.
Course Similar Figures Warm Up Solve each proportion. b = y5y5 = p9p9 = m = 4. b = 10y = 8 p = 3 m = 52.
1 Algorithms Practice Topics In-Class Project: Tip Calculator In-Class Project: Drawing a Rectangle.
Automated Conceptual Abstraction of Large Diagrams By Daniel Levy and Christina Christodoulakis December 2012 (2 days before the end of the world)
1 Mean Analysis. 2 Introduction l If we use sample mean (the mean of the sample) to approximate the population mean (the mean of the population), errors.
1 Friends and Neighbors on the Web Presentation for Web Information Retrieval Bruno Lepri.
SCALE FACTORS AND SIMILARITY Introducing scale factor.
A Scrollbar-based Visualization for Document Navigation Donald Byrd Proceedings of the 4 th ACM conference on Digital libraries, 元智資工所 系統實驗室 楊錫謦.
Speaker : Yu-Hui Chen Authors : Dinuka A. Soysa, Denis Guangyin Chen, Oscar C. Au, and Amine Bermak From : 2013 IEEE Symposium on Computational Intelligence.
Ad insertion at proxies to improve cache hit rates Amit Gupta and Geoffrey baehr, Sun Microsystems Laboratories 901 San Antonio Road Palo Alto,CA
John Lamping, Ramana Rao, Peter Porolli
Copyright © 2011, A Resource Allocation Mechanism of Data Center for Public Cloud Service 指導教授:王國禎 學生:連懷恩 國立交通大學網路工程研究所 行動計算與寬頻網路實驗室.
Mapping and Browsing the Web in a 2D Space Mao Lin Huang, Wei Lai, Yanchun Zhang. Tenth International Workshop on, 元智資工所 系統實驗室 楊錫謦 2000/7/12.
WebQuery: Searching and Visualizing the Web through Connectivity Jeromy Carriere, Nortel Rick Kazman, Software Engineering Institute 元智資工所 系統實驗室 楊錫謦 2000/1/5.
1 The Beauty of Mathematics For those who are mathematically inclined, there is often a definite aesthetic aspect to much of mathematics. Many mathematicians.
Similar Solids 12.7 Geometry. Similar Solids Two solids of the same type with equal ratios of corresponding linear measures (such as heights or radii)
Mingze Zhang, Mun Choon Chan and A. L. Ananda School of Computing
Identity and Search in Social Networks
Using Proportions with Similar Figures
Isometric and Oblique Pictorials
2 Understanding Variables and Solving Equations.
Finite element method.
Longest Common Subsequence (LCS)
Edit Distance 張智星 (Roger Jang)
Presentation transcript:

Fractal Approaches for Visualizing Huge Hierarchies Hideki Koike, Hirotaka Yoshihara Department of Communications and Systems University of Electro-Communications Chofu, Tokyo 182, Japan 元智資工所 系統實驗室 楊錫謦 1999/8/4

Outline Introduction Problems with Huge Hierarchies Fractal approach Discussion & Conclusion

Introduction Visualization systems for huge data structures(hierarchical) have a potential usefulness. It is meaningless to visualize such huge data with a current tree visualization framework because of the exponential explosion in size of a tree and the increase of visual element.

Problems with Huge Hierarchies(Cont.) : 1. Exponential explosion in size 。 For general tree layout The width of the tree increases exponentially corresponding to the depth of the tree. d n = N d n-1 (for normal tree layout) d n : the width which is necessary to draw all subtree under the tree at depth n. N : the branch degree of a tree.

Problems with Huge Hierarchies(cont.) : to generalize: d n = d 1 N n-1 The width increases exponentially. 。 For Cone Tree

Problems with Huge Hierarchies(cont.) : r n-1 / ( r n - r n-1 ) = sin½θ & θ= 2π/ N to generalize: r n = r 1 [ / sin(π/ N) ] n-1 The size also increases exponentially.

Problems with Huge Hierarchies(cont.) : 2. Scrolling is useless Why scrolling? -- The exponential explosion in size causes overflow off the screen. Why is scrolling not practical use? -- When the users focus on a node, they can’t see the minimum information in a scrollable window. -- It’s hard to trace links with scrolling to see a node’s neighbors.

Problems with Huge Hierarchies(cont.) : 3. Increase of Visual Elements It should be also considered how the increase of visual elements affects users’ cognitive load as well as the system’s response time.

Fractal approach : A word represents a wide variety of self- similar objects. 1. Fractal Views There exists a relation: D = - log r x N x between the branching factor, N, and the scale factor, r, at node x.

Fractal approach(cont.) : Fractal views calculate the degree of importance which is called fractal value ( Fv ) of each node, and decide which nodes should be displayed or erased based on these fractal values. The Fv x is calculated with the following equation. Fv focus = 1 Fv child-of-x = r x * Fv x (r x = CN x -1/D ) C is a constant value which satisfies 0 < C ≦ 1. For simplify calculation, C is set to 1 in our system.

Fractal approach(cont.) : The most important property of fractal view is the ability to control the amount of data displayed. The Fv x is calculated with the following equation. Fv focus = 1 Fv child-of-x = r x * Fv x (r x = CN x -1/D ) C is a constant value which satisfies 0 < C ≦ 1. For simplify calculation, C is set to 1 in our system.

Fractal approach(cont.) : The propagated value, Fv, of node at depth n is: Fv = r n = N -n/D ===> M = ( Fv -D - 1/N ) / ( 1 - 1/N ) ===> The number of nodes which have a value greater or equal to the threshold is nearly constant without a relation to the branching factor. In fisheye views, the degree of importance (DOI) of each node is calculated, with a distance from the root (-API) and a distance from the focus (D). DOI = API - D

Fractal approach(cont.) : 2. Fractal Tree Layout Using the self-similarity characteristic of a fractal, it is possible to standardize the view at every level of the tree. ==> In a fractal tree, when a part of the subtree is magnified, a similar view appears as well.

Fractal approach(cont.) : In the fractal tree layout, each node size is proportional to its fractal value.Thus, the number of recognizable nodes is almost constant. 3. Fractal Pruning All the displayed nodes, whether they are recognizable or not, reduce the system’s response time, these unrecognizable nodes as well as the nodes outside the viewing area should be erased.

Discussion & Conclusion: 。 Unsolved problem: Strictly speaking, this framework is not a 3D layout algorithm, because it does not check the overlap of cones. Solution: 1. A smaller fractal dimension minimizes the overlap of cones. 2. Self-avoidance -- the angle of each branch is decided independently.

Discussion & Conclusion(Cont.):

Discussion & Conclusion: 。 Conclusion 1. The geometrical characteristic of a fractal, self-similarity, madeit possible to visually interact with a huge tree in the same manner at every level of the tree. 2. The fractal dimension made it possible to keep the number of displayed nodes nearly constant. 3. Many problems remain unsolved.