Semantic Analysis and Complex Networks as Conjugated Techniques

Supporting Decision Making

Pedro Ivo Lancellotta, Victor Ströele, Regina Braga, José Maria N. David and Fernanda Campos

Post-Graduate Program in Computer Science, Federal University of Juiz de Fora, Juiz de Fora, Brazil

Keywords: Business Intelligence, Complex Networks, Data Visualization, Decision Making, InfoVis, Ontology.

Abstract: The expansion scenario in information technology has led to the need for the analysis of huge amounts of data

to manage information for quick decision making. This paper presents an architecture for data visualization

that uses semantic and structural interpretation as conjugated techniques for data analysis in different domains,

through an interface which supports visualization strategies. A case study was carried out with a specialist in

real-world agricultural context using data from dairy cattle to answer our research question. The results

demonstrate the feasibility of the proposal.

1 INTRODUCTION

Technology is increasingly present in the daily

routine of companies and industries as devices and

tools for data manipulation. The great amount of data

makes manual analysis a complex activity. The main

objective of Business Intelligence (BI) is to support

decision making, reduce costs, optimize work

efficiency to improve process management (Popovič

et al., 2013). Among the techniques to support BI

decision-making are Data Mining, Panels, Reports

and Data Visualization (Ranjan, 2008).

The use of data visualization techniques, or

information visualization, helps the cognitive process

in data interpretation and knowledge discovery (Card

et al., 2009). These aims to simplify the way these

data are presented to managers as well as decision

makers in order to be able to observe trends and

implicit information in huge volumes of data. Efforts

are made to create new mechanisms that optimize

information visualization techniques, such as the

identification of more appropriate visualization

techniques for specific datasets, and the reduction in

the amount of information displayed for a greater

abstraction of information, assisting user perception

(Jonker et al., 2013).

Among the efforts for the interpretation and

semantic data analysis, we can highlight the use of

ontologies, which have been applied in several

domains (Miah et al., 2007) (Zillner et al., 2008) (da

Silva & Cavalcanti, 2014) (Liu et al., 2016).

Ontologies add semantic value to the data, defining

explicit and implicit relationships among them. It can

define the knowledge related to a given domain of

data, allowing it to be processed computationally,

making possible the extraction of new information

and assignment of meaning to the data.

When defining an ontology for a domain, the

relationships between the domain interest individuals

and their characteristics, or in a simplified way, the

actors that interact in the domain of application are

identified. Thus, it is possible to delineate this

ontology in the form of a network, representing

individuals and their relationships (Heim et al., 2010).

Still according to (Heim et al., 2010) these networks

allow a more effective visualization of intrinsic

relationships of data, observing the connections

between the individuals that compose the network.

By characterizing the ontology in the form of a

network, it is possible to identify how data are related

and to apply complex network metrics, such as

centrality analysis, node degree, among other metrics

to identify individuals that stand out in the data set.

Several works use ontologies and complex

networks individually to provide interpretations of

the data, but few explore both together. This paper

explores the combination of ontology techniques and

complex networks using them together, to

complement the information on each one in an

automated or semi-automated form.

Ontologies are able to offer benefits to complex

networks, such as (i) Abstraction of network nodes

with the use of inferences (ii) Aggregation of

Lancellotta, P., Ströele, V., Braga, R., David, J. and Campos, F.

Semantic Analysis and Complex Networks as Conjugated Techniques Supporting Decision Making.

DOI: 10.5220/0006662001950202

In Proceedings of the 20th International Conference on Enterprise Information Systems (ICEIS 2018), pages 195-202

ISBN: 978-989-758-298-1

195

semantic values for the data network (iii)

Classification of individuals, identifying semantic

groups that can be analyzed separately. On the other

hand, the complex network can also benefit the

ontology, such as: (i) Identification of new groups of

individuals, based on the analysis of ontology entities,

enabling the creation of other groups with similar

characteristics identified in the network (ii)

Identification of new relationships between data,

automating the process of creating new logical rules

(iii) Structural analysis of the ontology and creation

of new properties of relationships between data, from

the use of complex network metrics.

The main objective of this paper is to present an

architecture that defines the necessary components

for the visualization of information and support for

decision making. The architecture uses ontologies

and complex networks in a conjugated form,

supporting visualization for decision-making.

To evaluate the technical feasibility of the

architecture, a visualization platform was developed

to support the components of the proposed

architecture. In addition, in order to evaluate whether

the proposal assists decision making in a real

environment, a case study was carried out for an

agricultural research company involving data

analysis of dairy cattle production.

The main contributions of this work are (i) the

proposal of a visualization architecture based on

semantic and structural data analysis (ii) Use of

ontologies and complex networks as conjugated

techniques in order to improve the interpretation of

strategic data (iii) Development of an ontology in the

agricultural context, related to dairy control, and (iv)

Development of a visualization platform to support

decision making.

The paper is organized as follows: Section 2

details some related works. In section 3, we present

the architecture. Section 4 presents a use case with

real data from dairy farms. Finally, section 5 presents

the conclusions and future work.

2 RELATED WORK

The main objective of this section is to identify works

that are related to the proposed approach in this paper,

as well as observing deficiencies found in the context

of data visualization. We reviewed works that use

information visualization, such as Business

Intelligence techniques, to support decision making.

We filtered the works that mainly use semantic and

complex network analysis as an approach,

contributing to the interpretation of information and

strategic analyzes for datasets in any domain.

Analyzing huge volumes of data is a challenge in

many research areas. Some research works, focus on

reducing the amount of data that will be available in

the visualization (Gu and Wang, 2011).

Ontologies have the ability to aggregate semantic

information to analyzed data, in addition to

incorporating several aids to the understanding of the

specification of a dataset (Sumalatha et al., 2008). A

recurrent use of the ontology is to consistently

characterize an application domain. It is possible to

represent actors that integrate it and its interrelations

creating a unified and consolidated concept of the

target domain (Jayaraman et al., 2013).

The use of complex networks allows a structural

investigation of the data, so that it is possible to

identify individuals who are more influential in the

network as well as their intrinsic relationships (Lopes

et al., 2010). Still according to (Lopes et al., 2010), in

order to identify important individuals within the

network, complex network metrics are used, such as

node degree, centrality, among others.

Many recent studies have made efforts to identify

algorithms and metrics for the identification of

relevant individuals and implicit relationships among

those individuals in complex networks, for example,

social networks (Silva et al., 2017) (Muniz et al.,

2017).

Some studies use ontologies and / or complex

network metrics to better visualize data and identify

important elements for analysis (Queiroz-Sousa et al.,

2013) (Ströele et al., 2017), and the creation of

visualization tools that behave in an automated or

semi-automated way (Sobral et al., 2016). However,

few studies use these techniques in conjunction with

data interpretation (Heim et al., 2010). Others only

use different views on data (Dörk et al., 2012) (Radics

et al., 2015) (Soklakova et al., 2016).

The main goal of this article is the integrated use

of ontology techniques and complex networks so that

one complements the information of the other. This

enables the development of a visualization platform

capable of inferring new information and rules about

a set of data analyzed in an automatic or semi-

automatic way.

3 CONCEPTUAL

ARCHITECTURE

Based on the literature review, we identified the need

to create a visualization mechanism to assist business

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

196

managers in making strategic decisions that lead to

higher quality and productivity. A conceptual

architecture that supports decision-making is

proposed which can self-feed and produce new

information. The general objective of the architecture

is to enable data visualization and interpretation from

a specific data source from the semantic and

structural point of view.

The conceptual model of the architecture is

presented in Figure 1. The architecture is divided into

layers, presented in detail in the following

subsections.

The operationalization of the architecture usage

occurs according to the following steps. Initially, (i)

data is collected from one or more repositories of the

same domain (ii) these data are organized, translated

into the ETL layer and then imported into the

ontology, in which some domain rules are defined

(iii) from the defined domain ontology, it is translated

as a network, showing the main relationships between

the main actors that constitute the application domain

(iv) finally, a complex network is modeled with the

individuals of the domain. In this network, the

relations can be similar to the connections obtained

from the ontology, or new connections between the

individuals are defined considering their intrinsic

characteristics. In this way, semantic analysis is done

for data abstraction in order to summarize, reduce the

data volume, highlight the most relevant data and

identify new rules or groupings.

3.1 ETL Layer

The ETL layer is responsible for extracting,

transforming and loading data for later processing.

Data repositories can come from several different

sources, but all of them refer to the same application

domain, or at least have data related to the domain.

In this vein, the architecture will import data and

load the ontology with the individuals, according to

the domain rules. The main functionality of the ETL

layer is the ability to interpret data from different

repositories to a consolidated and well-defined data

set through the ontology.

3.2 Processing Layer

The main purpose of the processing layer is to

organize and model the data so that it can be sent to

the visualization layer. Upon receiving data imported

by the ETL layer, the domain ontology is

consolidated with all individuals in the data set.

3.2.1 Domain Ontology

Semantic data interpretation is done through the

ontology, providing the benefits of reliability,

portability, reuse and maintenance.

The ontology is defined in order to consider the

interactions between the entities which are present in

the data set. Property chains are properties defined

through a chain of object properties, defining

inferences related to these data. The ontology has the

power to make the architecture more robust and

independent because it is able to interpret data, and

infer new knowledge. From this semantic layer, the

architecture covers different data domains. In

addition, these inferences support analysis based on

complex networks in order to reduce the number of

nodes to be represented through visualization

techniques, as well as to add semantic information to

the nodes and edges.

3.2.2 Ontology Network

In this layer, a network was modeled that represents

the relationships of the ontology and a heterogeneous

complex network.

The connection between elements that compose

the ontology naturally creates a network among the

actors who are inherent to the observed data domain.

The modeling and visualization of this network

allows some approaches for data analysis, such as, (i)

it illustrates to the decision maker the main links

between the elements that compose the data domain,

supporting the awareness and creation of new

relations from direct observation, and (ii) the use of

metrics (explained in section 3.2.4) to identify

relevant elements in the ontology.

3.2.3 Individual Complex Network

Nodes in a complex network represent individuals

and the connection between them represents different

measures that are defined by the user at run time.

This layer uses inferred data from the ontology

and reduces the amount of data to be presented to the

user. As a result, it identifies most relevant

individuals with complex network metrics.

Considering the semantic analyzes, the data

chosen by the user are related to the individuals

connected through the ontology. From the ontology,

relationships between data as well as their logical

rules are analyzed. Logical rules are responsible for

synthesizing information, so that individuals can

relate in a direct or indirect way.

Semantic Analysis and Complex Networks as Conjugated Techniques Supporting Decision Making

197

Figure 1: General overview of the architecture.

3.2.4 Complex Network Metrics

Complex network metrics can be used to perform the

analysis of highlighted nodes in the network. The

identification of more influential nodes is important

for a more specific analysis of an element, in order to

observe the causes of its prominence with respect to

the other individuals. Two metrics are used in the case

study in this work: closeness centrality and

betweenness centrality (Wasserman & Faust, 1994).

The closeness centrality measure represents the

average distance from the minimum paths of one

node to all other nodes in the network. Thus, the more

centered the node is, the closer it will be to other

nodes in the network. It is also possible to observe the

nodes with greater betweenness centrality, a metric

that totalizes how many minimum paths they pass

through a specific individual, allowing the

identification of entities that manage a greater range

of other entities and properties. Using these measures,

it is possible to identify new groups of individuals for

the creation of new semantic groups in the ontology,

that is, elaborating the semi-automated creation of

new logical rules.

3.2.5 Adapating Visualization

The processing layer will identify visualization

techniques according to the data types obtained from

the ontology and complex networks. It aims to

represent the data as clearly as possible to the decision

maker. Thus, the user can adapt the views to obtain

better understanding about the data. After choosing

the best visualization, the processing layer send this

information to visualization layer, which shows to the

user.

3.3 Visualization Layer

After the interpretation of the data by the ontology

and the structural analysis of the complex networks,

there is an interface layer in which the user makes the

choice of the data that is pertinent to his/her analysis.

He/she also defines the measurements of the data

he/she intends to observe, such as numerical values

that represent productivity, some attribute that

suggests accounting for something, or any other

attribute that is measurable.

4 EVALUATION

This section presents an initial evaluation of the

proposed architecture, which was made from the first

version of the information system for decision-

making support. The objective of this evaluation is to

answer the research question considering the use of

the proposed solution in a real-world context.

The evaluation method adopted in this paper is

based on a case study carried out in the context of an

agricultural research company with real data

extracted from dairy cattle production. A case study

is an empirical investigation that relies on multiple

sources of evidence to investigate an instance (or a

small number of cases) of a contemporary

phenomenon within its actual context, especially

when the boundary between phenomenon and context

cannot be clearly identified (Runeson et al., 2012).

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

198

4.1 Case Study Definition and Planning

A survey reported by MilkPoint (https://

www.milkpoint.com.br) highlights the managerial

aspect as the major limiting factor in milk production.

There is a lack of approaches that assist the managers

in the visualization of data on this field (Miah et al.,

2007).

The evaluation scope was defined based on the

GQM method (Basili et al., 1994) with the "purpose

of characterizing the effect of semantic and structural

analyzes, as aggregated techniques, on data

visualization approaches in decision making from

the point of view of business managers and milk

institutions in the context of milk quality".

The research question to be evaluated in this study

is: How can the semantic and structural data analysis

as aggregated techniques, combined with

visualization techniques support decision-making?

To collect evidence and respond to the proposed

research question, direct observations and interviews

were used as data sources.

4.2 Carrying out the Case Study

This section describes the semantic and complex

network analysis obtained from the analyzes

performed in the developed system. The interpretable

elements, observed from the use of ontologies and

networks, as well as their objectives and main

characteristics are described. The ontology developed

for the context of dairy cattle is presented, detailing

some of its logical rules and how its use can aid

decision-making.

4.2.1 Designing the Domain Ontology

As far as we are aware, no work in the literature was

found that used ontologies that represent the process

of milk control and quality from dairy cattle.

Therefore, an ontology was created using Protégé

(protege.stanford.edu). The creation of this ontology

was done with the help of specialists in the area of

agricultural research to better detail the process and

their needs. The ontology created has a total of 30

classes and 41 logical rules. The classes of Institution,

RegionalCenter, Coordinator, Consultant and

Producer represent managerial aspects and classes,

such as: DairyControl, Production, AnimalBreed,

Tank and Herd approach the productive aspect.

Through property chains (OWL 2.0), it is possible

to obtain the Farms that are served by Institutions or

other Entities, in addition to the productions or other

information related to their respective Farms. In this

context, it is possible to associate semantic analysis

with managerial processes.

4.2.2 Data Analysis

Initially, a comparative analysis was performed

through tables in relation to managerial data. The

domain specialists observed the hierarchical

relationship between some entities (Regional Center,

Coordinators and Consultants). Without the use of

ontologies, only the inter-relationships between

individuals was identified. Due to the data set

volume, it is costly to perform joins between these

data in order to identify the attributes and their

associations, allowing specialists to analyze only

direct relationships. The use of ontologies is able to

fill this gap, so that from the inferences made, it is

possible to aggregate values to this information. From

the platform expert's action to obtain managerial

information, a table with the data obtained with the

logical rules of the ontology is presented. In this case,

the production values of all properties served by

management entities were associated. As a result, it

was possible to make previously complex

observations, such as identifying the production of

properties served by a specific Regional Center.

For the definition of levels of abstraction and

structural analysis a complex network was modeled

considering the relationships between the individuals.

The management network aims to visualize the mean,

total or other aggregations on the values related to the

measurement of the dairy controls of animals from

one or more herds of a farm. In this case study, the

sum of the milk production values of the animals

during the last dairy control was used. The values of

these measurements are assigned as the weight of the

network edges.

With the help of this network, the specialist was

able to analyze which farms produce more milk for a

Regional Center chosen by him/her according to

his/her needs of analysis and decision making

In the context of milk quality and the ontology

used, there are two main networks that make up the

system: management and milk production. The

analyzes made by the specialist which focused on the

study of the managerial network (Figure 2) were

composed of the following entities: Institution,

Regional Center, Coordinator, Consultant and Farm.

In the management network presented in Figure 2,

the following elements are shown: regional centers

(in purple), coordinators (in pink), consultants (in

yellow) and, finally, farms (in gray). It is possible to

observe that, with the increase in the number of

institutions, the visualization of the network is

Semantic Analysis and Complex Networks as Conjugated Techniques Supporting Decision Making

199

compromised according to the volume of data

presented to the user. The objective is to highlight the

entities that serve farms with a higher milk yield.

For this, the complex network metrics defined

above are used. It was observed that nodes with

greater closeness centrality are related to more Farms

or with more Entities. Thus, comparing the entities of

the same type (Coordinator, Consultant, Regional

Center, and Institution) those that are most influential

in the network were identified and presented to the

specialist. This is due to the fact that they deal with a

higher volume of milk production. The betweenness

centrality metric was used to identify the nodes with

the greatest influence on entity and / or property

management. Thus, when measuring the minimum

paths between regional centers and farms, it was

possible to identify which entities are present in these

paths, showing that this entity manages a greater

range of other entities and properties.

In Figure 2, the ontology action can be observed

in the abstraction of the managerial network. With

this abstraction, the manager is able to briefly

visualize the measures identified with the total

analysis of the complex network. In this case, the

edge between two nodes (a consultant and a

coordinator) was selected, and the total production of

farms consulted by this one is 304829.3 liters of milk.

Thus, it is possible to observe, through the flow

that runs in the edges which coordinators or entities

influence more, considering the productivity of a

certain regional center or any other entity that one

wishes to take as the central element of the network.

The reduction in the number of nodes represented on

the screen (Figure 2) facilitates the visualization by

the business managers without compromising the

information contained in the complex network.

Finally, this case study demonstrated the use of

the first version of the system for visualization and

decision making that covers semantic and structural

aspects of the data. All the actions carried out in the

system, as well as the analyzes and metrics adopted

in this case study, were made by a specialist in the

agricultural field.

In addition to the direct observations, an interview

was conducted. Among the answers given by the

specialist, after the use of the developed system, three

can be highlighted.

(I) Is the information clearly presented by the system,

helping in the decision-making process?

“These charts enabled me to do an immediate

analysis of the production of specific farms

comparing to a national scenario, presenting

information that supported me in the decision-making

process with the milk producers”.

(II) How do you evaluate the use of different visual

resources (graphs, tables and network) for decision-

making?

“The graphs are important, but the need to analyze

them together stands out”.

(III) Is it possible to analyze data from different

institutions and their contribution to national cattle

production?

Figure 2: Complex Network generated from available data.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

200

“Milk production presented in a network way allows

us to identify and to compare the different entities in

relation to its productivity. This is a good

contribution of the system for data analysis”.

Based on the first answer, there is evidence that

the information presented in the visualization artifacts

provides better awareness, so that dairy institutions

can help milk producers in the administration of their

herds and farms. According to the second answer,

there is evidence that the components used helped

decision making. However, it is still necessary to

integrate these components. The third answer showed

that the use of semantic analysis and complex

networks as conjugated techniques can help in

decision-making process.

In order to answer the research question, we

observed (i) the specialists' interactions with the

system to obtain information; (ii) the analyzes carried

out by them for decision-making in the context of

milk quality; (iii) and the answers given by them in

the interview at the end of the case study. The

specialist's answers go to what was observed during

the case study, pointing to the proposal feasibility.

Considering these points, it is possible to affirm

that there is evidence that the semantic and structural

analysis of the data, together with visualization

techniques are able to aid in the decision making,

adding more information to the data. However,

further case studies should be carried out for a wider

evaluation of the proposal.

As threats to the validity of this case study we

can cite the reduced number of participants. Through

the evaluation it was possible to state that these results

could not be generalized. However, it was possible to

identify situations in which similar results could be

obtained.

5 FINAL REMARKS

This article presented an architecture for an

information system to support decision making,

based on components that use complex networks and

ontologies as conjugated techniques to support the

decision-making process for business managers. The

use of the ontology together with the complex

networks provides a semantic and structural analysis

of the data, so that the visualization strategy brings

forward additional information from the data used. A

classical problem in analyzing complex networks is

the amount of information presented to users. In this

sense, the ontology collaborates by providing

abstractions in the network, so that it is possible to

omit some nodes without the loss of important

information for decision-making. Complex networks

also collaborate in the creation of new logical rules

for the previously defined ontology.

A first version of the system was developed, and

a case study was carried out with the purpose of

evaluating the architecture. The use of ontologies and

complex networks of conjugated form adds value in

decision making. The results demonstrate the

feasibility of the proposal, but this is only an initial

evaluation. As future work, we intend to expand the

developed ontology, to carry out a case study

considering a greater number of specialists and, in

addition, adopt new metrics to support the analyzes

related to the complex networks and identification of

new logical rules.

ACKNOWLEDGEMENTS

We would like to thank the CAPES, CNPQ,

FAPEMIG, UFJF and EMBRAPA.

REFERENCES

Basili, V.; Caldiera, G.; Rombach, H.: Goal Question

Metric Approach, Encyclopedia of Software

Engineering (1994), John Wiley & Sons, Inc., p. 528-

532.

Card, S., Mackinlay, J., and Shneiderman, B. (2009).

Information visualization. Human-computer

interaction: design issues, solutions, and applications,

181.

Da Silva, M.A.A. and Cavalcanti, M.C. (2014). Combining

ontology modules for scientiﬁc text annotation. Journal

of Information and Data Management, 5(3):238.

Dörk, M.; Carpendale, S.; Williamson, C. (2012)

Visualizing explicit and implicit relations of complex

information spaces. Information Visualization, v. 11, n.

1, p. 5–21.

Gu, Y. and Wang, C. (2011). Transgraph: Hierarchical

exploration of transition relationships in time-varying

volumetric data. IEEE Transactions on Visualization

and Computer Graphics, 17(12):2015–2024.

Heim, P., Lohmann, S., and Stegemann, T. (2010).

Interactive relationship discovery via the semantic web.

The Semantic Web: Research and Applications, 303-

317.

Jayaraman, P. P., Gunasekera, K., Burstein, F., Haghighi,

P. D., Soetikno, H. S., and Zaslavsky, A. (2013). An

ontology-based framework for real-time collection and

visualization of mobile field triage data in mass

gatherings. In System Sciences (HICSS), 2013 46th

Hawaii International Conference on (pp. 146-155).

Semantic Analysis and Complex Networks as Conjugated Techniques Supporting Decision Making

201

Jonker, D., Langevin, S., Bozowsky, N., and Wright, W.

(2013, January). Aperture: An open web 2.0

visualization framework. In System Sciences (HICSS),

2013 46th Hawaii International Conference on (pp.

1485-1494).

Kawasaki, M., Kawamura, S., Tsukahara, M., Morita, S.,

Komiya, M., and Natsuga, M. (2008). Near-

infraredspectroscopicsensingsystemforon-

linemilkqualityassessment in a milking robot.

Computers and electronics in agriculture, 63(1):22–27.

Liu, Y., Wang, G., Jiang, Z., Qian, J., and Chen, Y. (2016,

October). An integrated decision support system for

emergency evacuation management. In Systems, Man,

and Cybernetics (SMC), 2016.

Lopes, G. R., Moro, M. M., Wives, L. K., and de Oliveira,

J. P. M. (2010). Cooperative Authorship Social

Network. In AMW.

Miah, S. J., Gammack, J., and Kerr, D. (2007). Ontology

development for context-sensitive decision support. In

Semantics, Knowledge and Grid, Third International

Conference on (pp. 475-478).

Muniz, C. P. M. T., Choren, R. and Goldschmidt, R. R.

(2017). Using a Time-based Relationship Weighting

Criterion to Improve Link Prediction in Social

Networks.

Petersen, Kai. "Measuring and predicting software

productivity: A systematic map and review."

Information and Software Technology 53.4 (2011):

317-343.

Popovič, A., Hackney, R., Coelho, P. S., and Jaklič, J.

(2012). Towards business intelligence systems success:

Effects of maturity and culture on analytical decision

making. Decision Support Systems, 54(1), 729-739.

Queiroz-Sousa, P. O., Salgado, A. C., and Pires, C. E.

(2013). A method for building personalized ontology

summaries. Journal of Information and Data

Management, 4(3):236.

Ranjan, J. (2008). Business justification with business

intelligence. Vine, 38(4), 461-475.

Radics, P., Polys, N., Neuman, S., and Lund, W. (2015).

Osnap! introducing the open semantic network analysis

platform. volume 9397.

Runeson, Per, et al. Case study research in software

engineering: Guidelines and examples. John Wiley &

Sons, 2012.

Silva, P. R., Dias, S. M., Brandao, W. C., Song, M. A., and

Zárate, L. E. (2017). Formal Concept Analysis Applied

to Professional Social Networks Analysis.

Sobral, T., Galvão, T., & Borges, J. (2016) VUMO:

Towards an ontology of urban mobility events for

supporting semi-automatic visualization tools. In

Intelligent Transportation Systems (ITSC), 2016 IEEE

19th International Conference on (pp. 1700-1705).

Soklakova, T., Ziarmand, A., and Osadchyieva, S. (2016,

October). Big data visualization in smart cyber

university. In East-West Design & Test Symposium

(EWDTS), 2016 IEEE (pp. 1-9).

Ströele, V., Campos, F., David, J. M. N., Braga, R.,

Abdalla, A., Lancellotta, P. I., Zimbrao, G., and Souza,

J. (2017). Data abstraction and centrality measures to

scientiﬁc social network analysis. Computer Supported

Cooperative Work in Design.

Sumalatha, M. R., Ravi, A., Aravind, M., & Prasanna, N.

(2008). Collective Intelligence in Distributed Systems

and Semantic Data Visualization. In Emerging Trends

in Engineering and Technology, 2008.

Wasserman, S. and Faust, K. (1994).

Socialnetworkanalysis: Methodsandapplications,

volume 8. Cambridge university press.

Zillner, S., Hauer, T., Rogulin, D., Tsymbal, A., Huber, M.,

and Solomonides, T. (2008). Semantic visualization of

patient information. In Computer-Based Medical

Systems, 2008.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

202