3D VISUALIZATION AND VIRTUAL REALITY FOR

VISUAL DATA MINING

A Survey

Zohra Ben Said, Fabrice Guillet

LINA, UMR 6241 CNRS, University of Nantes, Nantes, France

Paul Richard

LISA, EA 4094, University of Angers, Angers, France

Keywords:

Visualization techniques, Visual data mining, Virtual reality, Classiﬁcation.

Abstract:

Visual Data Mining (VDM) aims at an easier interpretation of data mining algorithm results through the

use of visualization techniques. During the last decade, many techniques of information visualization have

been proposed, allowing visualization of multidimensional data. Previously, ((Chi, 2000), (Herman et al.,

2000)) attempted to classify VDM techniques . However, these taxonomies do not take into account some

innovative techniques based on 3D visualization and virtual environments (VEs). In this paper, we propose

an exhaustive survey of recent techniques for VDM. These different techniques are detailed, classiﬁed and

compared according to the following criteria : graphical encoding, interaction techniques and applications.

Moreover, they are presented in tables together with graphical illustrations.

1 INTRODUCTION

Since the emergence of databases in the 60s, the vol-

ume of stored information grows exponentially each

year. In the 90s, this accumulation of information

in databases has motivated the development of a new

ﬁeld of research : Data Mining (DM) (Fayyad et al.,

1996). In many applications, such as network man-

agement (Tee et al., 2004), ﬁnance (Schreck et al.,

2007), seismic (Marroqun et al., 2008), users need

to explore relations in the data. These data sets are

often large and dynamic. In addition, understanding

data and tendencies is essential for users to make cor-

rect decisions. The extraction of useful tendencies

in data for the user (domain expert) constitutes the

main challenge of this research. The use of visualiza-

tion techniques proposed by VDM can improve the

readability of the results and offers signiﬁcant poten-

tial for interaction and exploration of large databases.

Given the number and variety of available visualiza-

tion techniques, it is a challenging activity for infor-

mation designers to ﬁnd out the methods, techniques

and corresponding tools available to visualize a par-

ticular type of information. The comparison of vi-

sualization techniques across different criteria is not

a trivial problem. Previously, ((Chi, 2000), (Herman

et al., 2000)) attempted to classify VDM techniques .

However, these taxonomies do not take into account

the latest approaches based on 3D and virtual real-

ity techniques. Visual Data Mining (VDM) is an ap-

proach to explorate data analysis and knowledge dis-

covery that is built on the extensive use of visual com-

puting. The basic goal is that large and incompre-

hensible amounts of data can be reduced to an easy

representation. This visual reprsentation can be eas-

ily understood and interpreted by a human. Accord-

ing to (Card et al., 1999), information visualization

allows the user to learn about data and relationships

among these data. The popularity of digital terrain

models (Simoff, 2001), based on the geographical

framework and CAD-based architectural models of

cities has demonstrated that multi-dimensional visu-

alization can provide a more efﬁcient way of explor-

ing large data sets. Some recent developments are ex-

tending VDM with algorithmic animation techniques,

multimedia support and virtual reality (VR) immer-

sive representations, aiming at involving decision-

makers in the mining and discovery process (Visual

Analytics). Decision-makers should be able to exam-

ine this massive, multi-dimensional, multi-source and

140

Ben Said Z., Guillet F. and Richard P. (2010).

3D VISUALIZATION AND VIRTUAL REALITY FOR VISUAL DATA MINING - A Survey.

In Proceedings of the International Conference on Imaging Theory and Applications and International Conference on Information Visualization Theory

and Applications, pages 140-145

DOI: 10.5220/0002850801400145

 SciTePress

time-varying information stream to make effective de-

cisions in time-critical situations (Keim et al., 2008).

Therefore, the success of VDM methods depend on

the development of adequate interaction and visual-

ization techniques.

Main Contributions. In this paper we purpose

• A recent review of 18 visualization techniques ac-

companied with graphic illustrations.

• A classiﬁcation of these techniques across 5

groups : Focus + context, 3D tree, virtual world,

3D scatterplot and dynamic graph.

• A comparaison of each group of techniques across

5 criteria : application, graphical encoding, inter-

action technique, advantages and drawbacks.

This paper is organized as follows. In section 2, we

describe focus + context visualization techniques. In

section 3 we present visualization techniques based

on 3D virtual worlds. The paper ends with a conclu-

sion.

2 FOCUS + CONTEXT

VISUALIZATION TECHNIQUES

Originally, the method of focus + context visualiza-

tion (F + C), aimed to wider detailds description of

cetain parts of data (the point of interest, focus, etc),

while the rest of the data is reduced in size in order to

provide a guidance to the users. The best techniques

F + C known, are the techniques of distortion: ﬁsheye

proposed by (Furnas, 1986). In the technique bending

backwards, another variant of the F + C technique,

the overview of different objects is not readable, but,

miniature views of objects are index in order to help

the user to move directly to the information sought.

However, there are other methods that the distortion

of space. The viewing volume for example, proposes

to vary the opacity (Mroz and Hauser, 2001), (color

shades) and frequency to achieve F + C visualization

of 3D data. A detailed comparison of these techniques

is presented in Table 1.

3 VIRTUAL WORLDS

VISUALIZATION TECHNIQUES

The virtual worlds (sometimes called cyber-spaces)

are another important trend in 3D information vi-

sualization. Virtual worlds for VDM are gener-

ally based either on the information galaxy metaphor

(Krohn, 1996) or the information landscape metaphor

(Robertson et al., 1998). The difference between the

two metaphors is that in information landscape, ele-

vation of objects is not used to represent information

(objects are placed on a horizontal ﬂoor). The speci-

ﬁcity of virtual worlds is that they provide to the user

some real-time 3D intuitive interaction and/or naviga-

tion techniques (control of the view point). A detailed

comparison of these approaches is presented in Table

(a): ARVis (b): sv3d

Figure 1: Illustrations of virtual worlds visualization tech-

niques.

3.1 3D Trees Visualization Techniques

Trees are information visualization techniques based

on hierarchical organization of the data. This ap-

proach ﬁnds many applications in graph visualiza-

tion. Indeed, 3D tree was designed to display a larger

number of nodes than those in 2D representations

(TreeMap (Johnson and Shneiderman, 1991)). The

conical trees are one of the best examples of this ap-

proach. They were introduced by (Robertson et al.,

1991) for visualizing large hierarchical structures in a

more intuitive way. 3D trees may be displayed verti-

cally (ConeTrees) or horizontally (CamTrees). Some

botanical approaches were proposed by (Ham and

Wijk, 2003) and (van de Wetering Kleiberg and van

Wijk, 2001).

3.2 3D Scatterplots

The 3D scatterplot visualization technique is one of

the most common representations in 3D scientiﬁc in-

formation visualization. It is based on the informa-

tion galaxy metaphor. The main innovation compared

to 2D visualization techniques is the use of volume

rendering that is a conventional technique in scien-

tiﬁc visualization (especially medical imaging). The

3D rendering techniques use voxels (instead of pix-

els) to represent a certain density of the data. This

technique has been adapted by (Becker, 1997), mak-

ing the opacity of each voxel a function of the density

of points.

3D VISUALIZATION AND VIRTUAL REALITY FOR VISUAL DATA MINING - A Survey

141

Table 1: Comparison of Focus + Context Techniques.

Visualization Applications Graphical Interaction Advantages Drawbacks

system encoding technique

Visualization Fish-eye

-CbVAR (Couturier

et al., 2007) Figure 2(a)

-Visualization

of association

rules

-2D :

context

-3D : focus

-Selection

-Zoom

-The display context help

orientation

-Displaying data in a

cluster

-Dynamic Tuning

-Few

parameters

displayed

- (Wang et al., 2008)

Figure 2(b)

-3D shape -Enlarge the

focal region

-Selection

-Zoom

-Deforming the non focal

region without

perceivable distortion

-Constraints in

the case where

there is not

enough space

Bending backwards

-3D-XV (Jacquemin

and Jardino, 2002)

Figure 2(c)

-Linear

structures

-Focus area

in the

center of

the screen

and near

data on the

sides

-Navigation

-Selection

-Different modes for

information accessibility

-Visualization

of sub-parts of

data at one

time

Linking and brushing

-Color

-WEAVE (Gresh et al.,

2000) Figure 2(d)

-SimVis (Doleisch

et al., 2005) Figure 2(e)

-Opacite

-RTVR (Mroz and

Hauser, 2001) Figure

2(f)

-The Magic Volume

Lens (Wang et al.,

2005) Figure 2(g)

-(Gtzelmann et al.,

2007)Figure 2(h)

-Frequence

(Elmqvist et al.,

2009)Figure 2(i)

-Medical

data,

scientiﬁc and

industrial

-Utilization

of colors

(SimVis,

WEAVE) ,

opacity

(RTVR)

and

frequency

emphasize

the

focussed

data parts

-Selection

-Feedback

(changing of

colors, etc.)

-Multiple linked views

-Immediate feedback

-Fast detection of

dependencies and

correlations

-No semantic

zoom

(a): cbVAR.

(b): (Wang et al., 2008).

(c): 3D-XV. (d) : WEAVE.

(e) : SimVis.

(g) : The magic volume lens.

(h) : (Gtzelmann et al., 2007).

(i) : (Elmqvist et al., 2009).

Figure 2: Illustrations of Focus + contexte Visualization Techniques.

IVAPP 2010 - International Conference on Information Visualization Theory and Applications

142

Table 2: Comparison of virtual worlds approaches.

Visualization Applications Graphical Interaction Advantages Drawbacks

system encoding techniques

Virtual worlds

-ARVis (Blanchard

et al., 2007) Figure

1(a)

-Visualization of

association rules

-Size of a cone, of a

sphere, their colors,

position of objects

on the arena

-Navigation

-Selection

-Zoom

-Order by set of

rules

-Navigation

according to

neighbor relation

-No hierarchical

representation

-Source Viewer 3D

(sv3D)(Maletic

et al., 2003) Figure

1(b)

-Visualization of

ﬁle structures

-Each code ﬁle is

represented by a

container.

-Color : type of the

control structure

-Navigation

-Zoom

-Selection

-Filtering

-History

-Screen shots

-Free

-The cylinder

position in the

container does not

represent any

variable

-No relations

between classes or

ﬁles

-No hierarchical

representation

3D trees

-SUMO (Buntain,

2008) Figure 3(a)

-OntoSphere3D

(Bosca et al., 2007)

Figure 3(b)

-Visualization of

ontologies

-Atom : concept

-Size of atom :

number of

documents

associated to the

concept

-A cluster :

concepts having

shared documents

-Zoom

-Navigation

-Easily interpretable -One hand

interaction

3D Scatterplots

-3D Scatter Plot

(VR) (Bovbjerg

et al., 2003) Figure

4(a)

-Visualization of

large data sets

-Different

colors/textures to

distinguish objects

and clusters

-The graphical

variables are :

position, shape,

size, color, sound

and texture

-Navigation

-Selection

-Zoom

-Use of sounds -Limited number of

graphical variables

-Not very efﬁcient

-VRMiner(VR)

(Azzag et al., 2005)

Figure 4(b)

- Visualization of

multimedia data

-Color, texture,

position, shape and

sound

-Zoom

-Navigation

-Selection

-Synthetic audio

-Use of VR

techniques

-Visualzation of

large images

-Low cost

-Limited number of

graphical variables

Dynamical graphs

-PEx (Paulovich

et al., 2007) Figure

-Visualization of

multidimensional

data

-3D projection of

multidimensional

data

-Color coding of

apparition

frequency

-Research

-Selection

-Personalization

-Filtering

-Free

-Visualization of

both structured and

non structured data

-No detail on

demand

(a): SUMO. (b): OntoSphere3D.

Figure 3: 3D trees visualization techniques.

3D VISUALIZATION AND VIRTUAL REALITY FOR VISUAL DATA MINING - A Survey

143

(a): 3DVDM. (b): VRMiner.

Figure 4: Examples of 3D Scatterplots.

3.3 Dynamic Graphs

Another technique based on the information galaxy

metaphor make use of dynamic graphs. Dynamic

graphs enable self organization data sets in the visu-

alization area. This approach is mainly used for the

visualization of hypertext or social networks. In this

context, a better approach is to apply a force system to

the nodes and links in order to ﬁnd a minimum energy

state of the system (or steady state) and determine the

position of the nodes.

PEx :(Paulovich et al., 2007).

Figure 5: Example of Dynamic graphs.

4 CONCLUSIONS

VDM aims at an easier interpretation of data mining

algorithm results through the use of intuitive and in-

teractive visualization techniques. In this paper we

proposed a recent review of 18 visualization tech-

niques accompanied with graphical illustrations. This

techniques are compared across 5 critera : applica-

tion, graphical encoding, interaction techniques, ad-

vantages and drawbacks. Even if, the main result is

that information visualization is indeed in great part

of application ﬁelds , this study shows that there is a

lack of interaction techniques. The main techniques

proposed, by most visualization techniques, are ba-

sic techniques like : zoom, selection, navigation. The

only system that offer a navigation through neighbor-

hood relations between data is ARvis. For an efﬁcient

data mining process, the user must be more involved

in the data mining process. Consequently, more so-

phisticated interaction techniques should be imple-

mented.

REFERENCES

Azzag, H., Picarougne, F., Guinot, C., and Venturini, G.

(2005). Vrminer: a tool for multimedia databases

mining with virtual reality. Processing and Managing

Complex Data for Decision Support, pages 318–339.

Becker, B. (1997). Volume rendering for relational data.

IEEE Symposium on Information Visualization.

Blanchard, J., Pinaud, B., Kuntz, P., and Guillet, F. (2007).

Visual analytics: A 2d-3d visualization support for

human-centered rule mining. Computers and Graph-

ics, 31(3):350–360.

Bosca, A., Bonino, D., Comerio, M., Grega, S., and Corno,

F. (2007). A reusable 3d visualization component for

the semantic web. In Web3D ’07: Proceedings of the

twelfth international conference on 3D web technol-

ogy, pages 89–96. ACM Press.

Bovbjerg, S., Granum, E., Nagel, H. R., and Vittrup,

M. (2003). Using dynamic soundscapes to sup-

port visual data mining in vr. In Simeon J. Simoff,

Monique Noirhomme-Fraiture, M. H. B. and Ankerst,

M. I., editors, Third International Workshop on Visual

Data Mining in conjunction with ICDM 2003 - The

Third IEEE International Conference on Data Min-

ing, pages 167–182.

Buntain, C. (2008). 3d ontology visualization in seman-

tic search. In Proceedings of the 46th Annual South-

east Regional Conference on ACM Southeast Regional

Conference, pages 204–208. ACM Press.

Card, S. K., Mackinlay, J. D., and Schneiderman, B. (1999).

Readings in information visualization : using vision to

think. Morgan Kaufmann publishers, San Francisco

CA, ETATS-UNIS (Monographie).

Chi, E. H. (2000). A taxonomy of visualization techniques

using the data state reference model. In INFOVIS ’00:

Proceedings of the IEEE Symposium on Information

Vizualization, pages 69–75. IEEE Computer Society

Press.

Couturier, O., Hamrouni, T., Yahia, S. B., and Nguifo, E. M.

(2007). A scalable association rule visualization to-

wards displaying large amounts of knowledge. In IV

’07: Proceedings of the 11th International Confer-

ence Information Visualization, pages 657–663. IEEE

Computer Society Press.

Doleisch, H., Mayer, M., Gasser, M., Priesching, P., and

Hauser, H. (2005). Interactive feature speciﬁcation for

simulation data on time-varying grids. In Conference

on Simulation and Visualization, pages 291–304. SCS

Publishing House e.V.

Elmqvist, N., Assarsson, U., and Tsigas, P. (2009). Dy-

namic transparency for 3d visualization : Design and

evaluation. The International Journal of Virtual Real-

ity, 8(1):75–88.

IVAPP 2010 - International Conference on Information Visualization Theory and Applications

144

Fayyad, U. M., Piatetsky-Shapiro, G., Smyth, P., and Uthu-

rusamy, R. (1996). Advances in knowledge discovery

and data mining. American Association for Artiﬁcial

Intelligence.

Furnas, G. W. (1986). Generalized ﬁsheye views. In

CHI ’86: Proceedings of the SIGCHI Conference on

Human Factors in Computing Systems, pages 16–23.

ACM Press.

Gresh, D. L., Rogowitz, B. E., Winslow, R. L., Scollan,

D. F., and Yung, C. K. (2000). Weave: a system for

visually linking 3-d and statistical visualizations, ap-

plied to cardiac simulation and measurement data. In

Proceedings of the conference on Visualization ’00,

pages 489–492. IEEE Computer Society Press.

Gtzelmann, T., Hartmann, K., Nrnberger, A., and Strothotte,

T. (2007). 3d spatial data mining on document sets for

the discovery of failure causes in complex technical

devices. In GRAPP’07 : Proceedings of the Second

International Conference on Computer Graphics The-

ory and Applications, pages 137–145. INSTICC - In-

stitute for Systems and Technologies of Information,

Control and Communication.

Ham, F. V. and Wijk, J. V. (2003). Beamtrees: compact

visualization of large hierarchies. Information Visual-

ization, 2(1):93–100.

Herman, I., Melancon, G., and Marshall, M. S. (2000).

Graph visualization and navigation in information vi-

sualization: A survey. IEEE Transactions on Visual-

ization and Computer Graphics, 6(1):24–43.

Jacquemin, C. and Jardino, M. (2002). Une interface

3d multi-

echelle pour la visualisation et la naviga-

tion dans de grands documents xml. In IHM ’02:

Proceedings of the 14th French-speaking conference

on Human-computer interaction (Conf

erence Fran-

cophone sur l’Interaction Homme-Machine), pages

263–266. ACM Press.

Johnson, B. and Shneiderman, B. (1991). Tree-maps: a

space-ﬁlling approach to the visualization of hierar-

chical information structures. In VIS ’91: Proceed-

ings of the 2nd conference on Visualization ’91, pages

284–291. IEEE Computer Society Press.

Keim, D. A., Mansmann, F., Schneidewind, J., Thomas, J.,

and Ziegler, H. (2008). Visual Analytics: Scope and

Challenges. Springer-Verlag, Berlin, Heidelberg.

Krohn, U. (1996). Vineta: navigation through virtual in-

formation spaces. In AVI’96 : Proceedings of the

workshop on Advanced visual interfaces, pages 49–

58. ACM Press.

Maletic, J. I., Marcus, A., and Feng, L. (2003). Source

viewer 3d (sv3d): a framework for software visualiza-

tion. In ICSE’03 : Proceedings of 25th ACM/IEEE

International Conference on Software Engineering,

pages 812–813. IEEE Computer Society Press.

Marroqun, V. D., Brault, J. J., and Hart, B. S. (2008). A vi-

sual data-mining methodology for seismic facies anal-

ysis: Part 2 - application to 3d seismic data. GEO-

PHYSICS, 74(1):13–23.

Mroz, L. and Hauser, H. (2001). Rtvr: a ﬂexible java li-

brary for interactive volume rendering. In VISUAL-

IZATION’01 : Proceedings of the Conference on Vi-

sualization, pages 279–286. IEEE Computer Society

Press.

Paulovich, F. V., Oliveira, M. C. F., and Minghim, R.

(2007). The projection explorer: A ﬂexible tool for

projection-based multidimensional visualization. In

SIBGRAPI’07 : Proceedings of the Brazilian Sympo-

sium on Computer Graphics and Image Processing,

pages 27–36. IEEE Computer Society Press.

Robertson, G., Czerwinski, M., Larson, K., Robbins, D. C.,

Thiel, D., and van Dantzich, M. (1998). Data moun-

tain: using spatial memory for document manage-

ment. In Proceedings of the 11th annual ACM sym-

posium on User interface software and technology,

pages 153–162. ACM Press.

Robertson, G. G., Mackinlay, J. D., and Card, S. K. (1991).

Cone trees: animated 3d visualizations of hierarchical

information. In CHI ’91 : Proceedings of the SIGCHI

conference on Human factors in computing systems:

Reaching through technology, pages 189–194. ACM

Press.

Schreck, T., Teku

sov

a, T., Kohlhammer, J., and Fellner, D.

(2007). Trajectory-based visual analysis of large ﬁ-

nancial time series data. ACM SIGKDD Explorations

Newsletter, 9(2):30–37.

Simoff, S. (2001). Towards the development of environ-

ments for designing visualisation support for visual

data mining. In Proceedings International Work-

shop on Visual Data Mining, pages 93–106. Simeon

J. Simoff, Monique Noirhomme-Fraiture, Michael H.

Bhlen and Mihael I. Ankerst.

Tee, S., J, T. T., kelly Kwan-liu Ma, J., and Wu, S. F.

(2004). Visual data analysis for detecting ﬂaws and

intruders in computer network systems. IEEE Com-

puter Graphics and Applications, special issue on Vi-

sual Analytics, 24(5):27–25.

van de Wetering Kleiberg, E. and van Wijk, J. (2001).

Botanical visualization of huge hierarchies. In NFO-

VIS ’01: Proceedings of the IEEE Symposium on

Information Visualization 2001, pages 87–94. IEEE

Computer Society.

Wang, L., Zhao, Y., Mueller, K., and Kaufman, A. (2005).

The magic volume lens: An interactive focus+context

technique for volume rendering. In VIS 05 : Proced-

ing of 16th IEEE Visualization, pages 367–374. IEEE

Computer Society Press.

Wang, Y.-S., Lee, T.-Y., and Tai, C.-L. (2008). Fo-

cus+context visualization with distortion minimiza-

tion. IEEE Transactions on Visualization and Com-

puter Graphics, 14(6):1731–1738.

3D VISUALIZATION AND VIRTUAL REALITY FOR VISUAL DATA MINING - A Survey

145