GUIDELINES FOR THE CHOICE OF VISUALIZATION

TECHNIQUES APPLIED IN THE PROCESS OF KNOWLEDGE

EXTRACTION

Juliana Keiko Yamaguchi, Maria Madalena Dias and Clélia Franco

Department of Informatic, State University of Maringá, Av. Colombo - 5.790, Maringá, Brazil

Keywords: Data visualization, Visualization techniques, Knowledge discovery, Data visualization parameters.

Abstract: Visualization techniques are tools that can improve analyst's insight into the results of knowledge discovery

process or to directly explore and analyze data. They allows analysts to interact with the graphical

representation to get new knowledge. The choice of visualization techniques must follow some criteria to

guarantee a consistent data representation. This paper presents a study based on Grounded Theory that

indicates parameters for select visualization techniques, which are: data type, task type, data volume, data

dimension and position of the attributes in the display. These parameters are analyzed in the context of

visualization technique categories: standard 1D - 3D graphics, iconographic techniques, geometric

techniques, pixel-oriented techniques and graph-based or hierarchical techniques. The analysis over the

association among these parameters and visualization techniques culminated in guidelines establishment to

choose the most appropriate techniques according to the data characteristics and the objective of the

knowledge discovery process.

1 INTRODUCTION

Represent the gained information in a visual way is a

solution for facilitate the understanding of analized

data. Thus, visualization techniques can be

integrated into the process of KDD, so much to

preview data to be analyzed or help in understanding

the results of data mining, so much to understand the

partial results of the iterations in the process of

extracting knowledge (Ankerst, 2001).

However, exploration and analysis of data using

visualization techniques directly applied to data can

bring useful and new knowledge, enough to exempt

other data mining techniques. Furthermore,

visualization is a powerful tool for conveying ideas,

due to vision plays an important role in human

cognition (Nascimento and Ferreira, 2005).

When visualization techniques are chosen to data

analysis, some criteria must be considered so that the

graphical representation really helps in

understanding data. First of all, it should be

observed relevant characteristics of the data, such as

data type, dimensionality (number of attributes) and

volume.

Tasks that users can perform during data

exploration may also be another factor in this

decision. Basically, the following tasks are the most

common mentioned in the literature: data overview,

veriﬁcation of correlation among attributes,

identiﬁcation of new rules or patterns, cluster

analysis and outliers detection. Furthermore,

depending on the visualization technique used,

positioning of the attributes in the graph can be

signiﬁcant in interpreting the behavior of data.

This paper presents a study whose research

methodology was based on Grounded Theory to

establish guidelines for choosing most suitable

visualization techniques according to the

characteristics of analyzed data.

So, in next section Grounded Theory

methodology is brieﬂy presented. Following, the

items: data type, dimensionality, volume, task type

and positioning of attributes in the graphic are the

named parameters identiﬁed through this

methodology and they are described in sequence.

Next, an analysis on the association of each

parameter to different types of visualization

techniques is discussed. After this, general

guidelines for choosing visualization techniques

according to the identiﬁed parameters are outlined.

183

Keiko Yamaguchi J., Madalena Dias M. and Franco C..

GUIDELINES FOR THE CHOICE OF VISUALIZATION TECHNIQUES APPLIED IN THE PROCESS OF KNOWLEDGE EXTRACTION.

DOI: 10.5220/0003469901830189

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 183-189

ISBN: 978-989-8425-53-9

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Finally, last section presents the conclusion of this

work.

2 RESEARCH METHODOLOGY

Grounded Theory (GT) was proposed in 1967 by

both sociologists, Glaser and Strauss, as an

alternative to traditional scientiﬁc method that it

normally consists of problem formulation, deﬁnition

and veriﬁcation of the hypothesis and conclusion.

GT, in turn, does not start from any hypothesis.

In this work was employed the following steps of

GT for the guidelines preparation for choosing

visualization techniques, based on GT versions of

the authors: Rodon and Pastor (2007) and

Orlikowski (1993), which are illustrated in Figure 1

and described as follows.

Figure 1: The Grounded Theory Process followed.

 Data collection - in this stage, the literature

was used as a data source, in which

information about visualization were

selected and analyzed in order to replace

original data could be obtained from

interviews, questionnaires and forms (Dick,

2005).

 Theoretical sampling - it is done alongside

with data collection, gathering all analyzed

data and comparing them with new

informations.

 Open coding - from theoretical sampling the

data are classiﬁed according to their

similarities. Each class is identiﬁed for a

code which, in this case, represents the

parameters to be considered to choose

visualization techniques. It was used the key

point coding method (Allan, 2003).

 Memoing - consists of analyst’s annotations

about encoding process and the analyzed

data, providing extra information for

construct theory.

 Axial coding - corresponding to arrange the

categories deﬁned in open coding and the

concepts connecting them for contextualize

the theory. Thus, in this step the guidelines

are deﬁned from the association between the

identiﬁed parameters and the categories of

visualization techniques.

 Diagramming - comprises constructing a

diagram to illustrate the concepts and

categories in order to facilitate the

understanding of the theory or phenomenon

that concludes the study.

 Theory generation - this work has as resulted

guidelines for choosing visualization

techniques based on parameters that

inﬂuence this choice according to the

characteristics of the data, being the

guidelines considered as the generated

theory in this step.

 Theoretical saturation - when new data will not

inﬂuence the organization and structure of

categories and concepts previously deﬁned,

it is an indication that the theoretical

saturation point was reached and,

consequently, the theoretical sampling

represents the scope of research.

The identiﬁed parameters through open coding

process are described in next.

3 GETTING THE PARAMETERS

The parameters to consider in select visualization

techniques emerged from open coding by analyzing

the literature related to visualization techniques, in

which we used the key points encoding method. The

result is illustrated in Table 1.

In the ﬁrst column of this table, the expressions

was taken from the related works, whose references

are in the next column, for each one was assigned

the concepts, described in the third column. Thus it

was possible to identify the parameters: data type,

task type, volume, dimensionality and position of the

attributes in the graph, which compose the aspects to

be considered in the decision to adopt visualization

techniques to represent data.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

184

Table 1: Key point coding method applied on collected data.

Key points References Code

Visualization techniques can be classified, among other criteria,

by data type

Shneiderman (1996)

Freitas et al. (2001)

Keim (2002)

Data type

Task type is one of the aspects considered in classification of

visualization techniques, which provides means of interation

between the analyst and the display

Shneiderman (1996)

Keim (2002)

Pillat et al. (2005)

Task type

Visualization techniques are subject to some limitations, such as

the amount of data that a particular technique can exhibit

Keim e Kriegel (1996)

Oliveira e Levkowitz (2003)

Rabelo et al. (2008)

Volume

Visualization techniques can also be classified according to the

number of attributes

Shneiderman (1996)

Grinstein et al. (2001)

Keim (2002)

Oliveira e Levkowitz (2003)

Dimensionality

In some category of visualization techniques, distribution form of

attributes on the chart can influence the interpretation about the

representation, such as correlation analysis, in which the relative

distance among the plotted attributes is relevant for observation

Ankerst (2001)

Oliveira e Levkowitz (2003)

Inselberg (2008)

Klippel et al. (2009)

Positioning of

attributes

4 GUIDELINES FOR SELECTION

OF VISUALIZATION

TECHNIQUES

Axial coding is the next step after the identiﬁcation

of the parameters. It was done based on analyzes

about the relationship between the parameters and

categories of visualization techniques, according to

the classiﬁcation suggested by Keim (2002), who

distinguishes five classes of techniques: (1) standard

1D-3D graphics; (2) iconographic techniques; (3)

geometric techniques; (4) pixel-oriented techniques;

and (5) based on graphs or hierarchical techniques.

Standard graphics are commonly used in statistic

to view an estimate of certainty about a hypothesis

or the frequency distribution of an attribute or to

view a data model. In iconographic techniques data

attributes are mapped into properties of an icon or

glyph, which vary depending on the values of

attributes. In geometric techniques,

multidimensional data are mapped into a two-

dimensional plane providing an overview of all

attributes. In pixel-oriented techniques, each value

of attribute is mapped to a pixel color and it is

placed on the display screen, divided into windows,

each corresponding to an attribute. In the end, they

are arranged according to different purposes (Keim,

2000). Data with a naturally structure of

relationships among its elements, hierarchical or

simple network, may be represented by hierarchical

or graph-based techniques.

It was chose some visualization techniques of

each category to illustrate the analysis of the

parameters. Among standard charts, it was selected:

the Histogram; the Box Plot; the Scatter Plot and the

Contour Plot. Among Icon-based: the Chernoff

Faces; the Star Glyphs and the Stick Figure. Among

Geometrically transformed displays: the Scatter Plot

Matrix and the Parallel Coordinates. Among Pixel-

oriented displays: the Query-dependent and the

Query-independent techniques. Among Graph-based

or Hierarchical: the Graph; the Cone Tree; the

Treemap; the Dimensional Stacking and the Mosaic

Plot. Next, each parameter is analysed with the

techniques mentioned above.

4.1 Analysis on the Data Type

Parameter

Techniques of standard 1D-3D category generally

represents from one to three attributes and are used

for analysis of quantitative data in most cases. All

charts considered in this class are able to display

quantitative data. To represent qualitative data,

alternative techniques are more limited. From the

selected techniques of this class only the Histogram

is able to plot qualitative data (Myatt, 2007).

Iconographic techniques are more appropriate for

quantitative data, because icon features vary with the

values of represented attributes. In Chernoff faces,

the shapes of each facial properties are changed; in

Star Glyph, the components of the star are modiﬁed;

in Stick Figure, the format of segments are different

according to the value of attributes.

GUIDELINES FOR THE CHOICE OF VISUALIZATION TECHNIQUES APPLIED IN THE PROCESS OF

KNOWLEDGE EXTRACTION

185

Geometric techniques are more ﬂexible, being

able to represent quantitative and qualitative data.

This applies to the Parallel Coordinates technique,

which can display attributes of these two data types.

Due to the scatterplot matrix is formed by a set of

scatterplots, it is more suitable for continuous

quantitative data.

In literature are found examples of usage of

pixel-oriented techniques on quantitative data.

Query-independent techniques were applied to

represent temporal data, and query-dependent

techniques are commonly used to represent

continuous quantitative data. Keim (2001) states

that they are not recommended for displaying

qualitative data.

Hierarchical or graph-based techniques are ideal

for displaying data when they have a structure of

relationships among themselves or with a structure

of hierarchy or simple network.

4.2 Analysis on the Task Type

Parameter

Generally, some techniques are better for certain

tasks than others. A task type execution depends if it

is implemented by the tool in use according to the

goals in improve the exploitation data activity.

Task type refers to activities that user or analyst

can perform according to goals in the use of a

graphical representation as noted in the literature

(Keim, 2002; Shneiderman, 1996; Pillat et al.,

2005). For practical purposes, the most common

tasks were considered in this work, such as:

 Overview data: view the whole data collection;

 Correlation among attributes: the degree of

relationship among variables can reveal

patterns of behaviour and trends;

 Identiﬁcation of rules, standards and important

characteristics;

 Clusters identiﬁcation: attributes with similar

behaviour;

Outliers detection: data set with atypical

behaviour in comparison for the rest of data.

Standard 1D-3D techniques serve, in general, to

view an estimate of certainty about a hypothesis or

the frequency distribution about an attribute, such as

the usage of histogram. This class also provides

graphs to make comparisons and data classiﬁcations

(in this case, can be used box plot), and also to

determine the correlation between attributes.

Different statistical graphs can be used in data

analysis, in order to discover patterns and structures

in data and identify outliers that can be observed, for

example, through using box plot or scatterplot.

Iconographic techniques represent each data

entry individually, allowing veriﬁcation of rules and

behaviour patterns of the data. Icons with similar

properties can be recognized and thus form groups

and it be analysed in particular. A representation

with a discrepant format if compared to the other

may characterize an outlier.

Geometric techniques provide a good overview

of the data, assigning no priorities to represent its

attributes. Furthermore, veriﬁcation of correlation

among them may be more discerning when using

techniques of this class, such as the scatterplot

matrix. This category of techniques also allows the

identiﬁcation of patterns, rules and behaviours and

may also detect outliers, characterized by behaviours

outside the common standard. The analyst may

choose to analyse a group of data that can be

detached from the tool in use but, in principle,

groups are not immediately identiﬁed by techniques

of this class.

Pixel-oriented techniques can be used in the

analysis of relationships among data attributes, to

find data cluster, so rules and patterns may be

identiﬁed through observing the correlations among

them.

Hierarchical techniques are useful for

exploitation of data arranged in a hierarchical or

simple relationship. Through techniques of this class

is possible to obtain an overview of the data

structure and analyse the relationship among the

elements. Techniques of this category also allow

grouping data, such as Treemaps (Shneiderman,

2006).

4.3 Analysis on the Volume and

Dimensionality Parameters

Implementations of visualization techniques must

take in consideration the limits of dimensionality

and volume of data to hold in way that the tool be

capable of providing a clear overview of data to the

analyst.

Standard graphics has low dimensional, because

they are intended to represent data with one to three

attributes. In addition, they support the view of a

small volume of data because, in general, they come

from statistical studies, resulting of a sample or of

percentages.

Iconographic techniques are able to handle a

larger number of attributes in comparison to the

standard graphics; however, the visualization

generated is best for a small amount of data due to

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

186

the space occupied by the icons in the screen. This is

the same statement found in (Rabelo, 2008), in

which the iconographic techniques evaluated (Star

glyphs and Chernoff Faces) were classiﬁed as low

scalability (support to display an amount of data).

Geometric techniques, in turn, may work with an

increased number of dimensions and volume when

compared to standard 1D-3D graphics and

iconographic techniques. But they are outweighed

by the pixel-oriented techniques for their capability

to represent the largest volume.

Hierarchical techniques or graph-based

techniques are usually used to represent the

relationship among data, regardless of

dimensionality, which can be high or low, but have

the same space constraints like that presented by

iconographic techniques, being the visualization

clearer if the amount data is not bulky.

However, visualization tools can offer features

like zoom, select, ﬁlter, among others, to improve

the interactivity with the visualization, mitigating the

limitations of each technique.

4.4 Analysis on the Positioning of

Attributes Parameter

Although it is not a parameter directly linked to the

characteristics of data, it is an important factor in

visual data exploration for some techniques as,

among others, Treemaps (Shneiderman, 2006),

Mosaic Plots (Hofmann, 2008), Dimensional

Stacking (LeBlanc et al., 1990). This parameter

depends on the technique or tool used to generate the

visualization, which should allow the change of the

positions of the attributes in the graph, producing

different views that can reveal new patterns.

In general, for 1D-3D standard graphics,

positioning of attribute do not change the

interpretation of results due to the low

dimensionality of the data that might be represented.

Moreover, the goal of using techniques of this class

is to analyze the behavior of a given attribute, or the

correlation among two or three attributes.

Stick Figures is an example in which the position

of the attributes can inﬂuence the visual data

exploration according to the icon type used, derived

from the variation of the mapping of data attributes

into icon properties (Pickett and Grinstein, 1988).

Chernoff faces, in turn, have a ﬁxed structure for

its icon, since it corresponds to the human face

characteristics and thus, the change of the

positioning of attributes is not a relevant aspect for

this technique.

But there are studies about which icon properties

may be more representative for the interpretation of

results, such as the eyes size and the shape of the

face are aspects that draw attention (Morris et al.,

2000; Lee et al., 2003). Likewise it is for Star Glyph

technique, for which once established the order of

the best mapping of attributes (Peng et al., 2004;

Klippel et al., 2009), it remains the same for all the

icons representing a record data per star.

In the works of (Inselberg, 2008) and (Wegman,

1990), it is explained how the position of the

attributes in the graph may inﬂuence the correlation

detection in Parallel Coordinates. Scatteplot Matrix

is, on the other hand, composed of a set of

scatterplots, for this reason nor is inﬂuenced by the

change of attributes positioning, since their main

objective is to evaluate the correlation between

attributes.

Keim (2000) presents techniques for the

placement of pixels on the display, which can

inﬂuence the interpretation of the visualization to

identify patterns and relationships among the

represented attributes.

The query-independent technique, for example,

may have the pixels arranged by recursive pattern

technique. When using the query-dependent

technique, the pixels can be arranged in the window

using spiral technique (Keim, 1997).

Hierarchical techniques or graph-based

techniques are in general inﬂuenced by the attributes

positioning, due to its elements naturally hold a

relationship structure, therefore, the assignment of

variables in the graph should be made carefully,

especially when there is a hierarchy between the

elements. The exception is for the Cone Trees

technique, which represents a deﬁned tree structure

(as ﬁles and directories structures in a hard disk),

providing only interactive features such as animation

to navigate among the tree nodes (Cockburn and

McKenzie, 2000; Robertson et al., 1991).

5 CONCLUSIONS

The Grounded Theory provided a methodology for

the identiﬁcation of the parameters and guidelines

for choose visualization techniques, set forth through

the stages theoretical sampling, coding,

diagramming and formulation of the theory.

During the development of this work, ﬁve

parameters were identiﬁed: data type, task type,

volume and dimensionality of the data and position

of the attributes in the graph. Subsequently these

parameters were analyzed in relation to the

GUIDELINES FOR THE CHOICE OF VISUALIZATION TECHNIQUES APPLIED IN THE PROCESS OF

KNOWLEDGE EXTRACTION

187

categories of visualization techniques distinct among

1D-3D standard graphics, iconographic techniques,

geometric techniques, pixel-oriented techniques, and

hierarchical techniques or graph-based techniques.

Through analysis of relationship among the

parameters and the visualization techniques, it was

observed that each technique type have a certain

conﬁguration of parameters that reﬂect the

characteristics of data and the objectives of the use

of visualization.

Data type must be the ﬁrst parameter to be

considered. It is the type of data that determines

what kind of visualization technique can be a priori

used. Qualitative data, for example, will be hardly

understood if they were represented by a technique

developed to represent quantitative data and vice

versa. Furthermore, it was veriﬁed in this study that

there are more options for visualization techniques

to represent quantitative data than qualitative data.

The task type to be performed corresponds to the

goals of the analyst during the data exploration. In

literature are found classiﬁcations of visualization

techniques based on this parameter. For tasks related

to statistical analysis, for example, the graphics 1D-

3D may be sufﬁcient; for tasks of correlation

veriﬁcation may be used visualization techniques of

geometric category, and so on.

Both volume and dimensionality of data are

limiting factors for visualization techniques.

Although most of them supports multidimensional

data, usually these techniques differ in the ability to

display a certain amount of dimensionality and

volume of data.

This is the case of categories of techniques

iconographic, geometric and pixel-oriented.

However, other ways of interaction can be used

during the visual exploration to minimize these

limitations, for example, the functions of zooming,

selection and ﬁlter.

The positioning of the attributes is a factor more

dependent on visualization technique to be used and,

hence, on the tool that implements it. For some

techniques such as parallel coordinates and star

glyphs, positioning of attributes is important for

discovery new patterns or behaviors. In the case of

parallel coordinates, positioning of attributes

inﬂuences the way data are displayed in polygonal

lines; in star glyphs technique, the order of

distribution of attributes to the icon properties can

ease grouping task, considering that arranging them

in different orders may generate diverse icon

formats.

Besides the parameters, another important point

to consider is the analyst’s familiarity with the data

analyzed. This is what will awaken new interests or

stimulate the users curiosity during data exploration,

forming new hypotheses that can be veriﬁed by

means of visualizations, or simply comparing the

results generated by the graphical representations.

It should be noted that the guidelines were

established based on the strongest features identiﬁed

during the coding phase. This does not mean the

invalidation of the use of a visualization technique

for other purposes that differ from those established

by the guidelines.

Therefore, visualization is a beneﬁcial tool in

understanding the knowledge, which may be

achieved through data mining algorithms or by

visual exploration performed directly on the data.

ACKNOWLEDGEMENTS

This work was supported by the Fundação

Araucária.

REFERENCES

Allan, G. (2003). A critique of using grounded theory as a

research method. Electronic Journal of Business

Research Methods, 2(1):1–10.

Ankerst, M. (2001). Visual data mining with pixel-

oriented visualization techniques. In Proceedings of

the ACM SIGKDD Workshop on Visual Data Mining.

Citeseer.

Cockburn, A. and McKenzie, B. (2000). An evaluation of

cone trees. People and Computers, pages 425-436.

Dick, B. (2005). Grounded theory: a thumbnail sketch.

Freitas, C., Chubachi, O. M., Luzzardi, P. R. G., and Cava,

R. A. (2001). Introdução à visualização de

informações. Revista de Informática Teórica e

Aplicada, 8(2):143–158.

Grinstein, G., Trutschl, M., and Cvek, U. (2001). High

dimensional visualizations. In Proceedings of the 7

Data Mining Conference-KDD. Citeseer.

Hofmann, H. (2008). Mosaic plots and their variants. In

Chen, C., Härdle, W., and Unwin, A., editors,

Handbook of Data Visualization, pages 617–642.

Springer.

Inselberg, A. (2008). Parallel Coordinates: Visualization,

Exploration and Classiﬁcation of High-Dimensional

Data. In Chen, C., Härdle, W., and Unwin, A., editors,

Handbook of Data Visualization, pages 643–680.

Springer.

Keim, D. A. (1997). Visual techniques for exploring

databases.

Keim, D. A. (2000). Designing pixel-oriented visual-

ization techniques: Theory and applications.

Visualization and Computer Graphics, IEEE

Transactions on, 6(1):1–20.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

188

Keim, D. A. (2001). Visual exploration of large data sets.

Communications of the ACM, 44(8):38–44.

Keim, D. A. (2002). Information visualization and visual

data mining. IEEE Transactions on Visualization and

Computer Graphics, 8(1):1–8.

Keim, D. A. and Kriegel, H. (1996). Visualization

techniques for mining large databases: A comparison.

IEEE Trans. on Knowl. and Data Eng., 8(6):923–938.

Klippel, A., Hardisty, F., and Weaver, C. (2009).

Starplots: How shape characteristics inﬂuence

classiﬁcation tasks. Cartography and Geographic

Information Science, 36(2):149–163.

LeBlanc, J.,Ward,M. O., andWittels, N. (1990). Exploring

n-dimensional databases. In Proceedings of the 1st

conference on Visualization’90, page 237. IEEE

Computer Society Press.

Lee, M. D., Reilly, R. E., and Butavicius, M. E. (2003).

An empirical evaluation of chernoff faces, star glyphs,

and spatial visualizations for binary data. In

Proceedings of the Asia-Paciﬁc symposium on

Information visualisation, Volume 24, pages 1–10.

Australian Computer Society, Inc.

Morris, C. J., Ebert, D. S., and Rheingans, P. (2000).

Experimental analysis of the effectiveness of features

in chernoff faces. In Proc Spie Int Soc Opt Eng,

volume 3905, pages 12–17. Citeseer.

Myatt, G. J. (2007). Making sense of data: a practical

guide to exploratory data analysis and data mining.

Wiley-Blackwell.

Nascimento, H. A. D. and Ferreira, C. B. R. (2005).

Visualização de informações – uma abordagem

prática. In XXV Congresso da Sociedade Brasileira de

Computação, XXIV JAI, São Leopoldo, RS, Brazil.

UNISINOS.

Oliveira, M. C. F. and Levkowitz, H. (2003). From visual

data exploration to visual data mining: A survey. IEEE

Transactions on Visualization and Computer

Graphics, 9(3):378–394.

Orlikowski, W. J. (1993). Case tools as organizational

change: investigating incremental and radical changes

in systems development. MIS quarterly, 17(3):309–

340.

Peng, W., Ward, M. O., and Rundensteiner, E. A. (2004).

Clutter reduction in multi-dimensional data

visualization using dimension reordering. In

Information Visualization, 2004. INFOVIS 2004. IEEE

Symposium on, pages 89–96. IEEE.

Pickett, R. M. and Grinstein, G. G. (1988). Iconographic

displays for visualizing multidimensional data. In

Proc. IEEE Conf. on Systems, Man and Cybernetics,

IEEE Press, Piscataway, NJ, volume 514, page 519.

Pillat, R. M., Valiati, E. R. A., and Freitas, C. M. D.

S.(2005). Experimental study on evaluation of

multidimensional information visualization

techniques. In Proceedings of the 2005 Latin

American conference on Human-computer interaction,

pages 20–30. ACM.

Rabelo, E., Dias, M., Franco, C., and Pacheco, R. C.

S.(2008). Information visualization: Which is the most

appropriate technique to represent data mining

results?In CIMCA ’08: Proceedings of the 2008

International Conference on Computational

Intelligence for Modelling Control & Automation,

pages 1228–1233, Vienna, Austria. IEEE Computer

Society.

Robertson, G. G., Mackinlay, J. D., and Card, S. K.

(1991). Cone trees: animated 3d visualizations of

hierarchical information. In Proceedings of the

SIGCHI conference on Human factors in computing

systems: Reaching through technology, pages 189–

194. ACM.

Rodon, J. and Pastor, J. (2007). Applying grounded theory

to study the implementation of an inter-organizational

information system. Electronic Journal of Business

Research Methods, 5(2):71–82.

Shneiderman, B. (1996). The eyes have it: A task by data

type taxonomy for information visualizations. In VL

’96: Proceedings of the 1996 IEEE Symposium on

Visual Languages, pages 336–343, Boulder, Colorado.

IEEE Computer Society.

Shneiderman, B. (2006). Discovering business intelligence

using treemap visualizations. Technical report, B-Eye:

Business Intelligence Network.

Wegman, E. J. (1990). Hyperdimensional data analysis

using parallel coordinates. Journal of the American

Statistical Association, 85(411):664–675.

GUIDELINES FOR THE CHOICE OF VISUALIZATION TECHNIQUES APPLIED IN THE PROCESS OF

KNOWLEDGE EXTRACTION

189