Adaptive System to Support Decision-making of Dairy Ecosystem in
Boyacá Department
Javier Antonio Ballesteros-Ricaurte
1,3
, Angela Carrillo-Ramos
2,3
,
Carlos Andrés Parra Acevedo
2,3
and Juan Erasmo Gómez Morantes
2,3
1
Escuela de Ingeniería de Sistemas y Computación, Universidad Pedagógica y Tecnológica de Colombia, Tunja, Colombia
2
Depto de Ingeniería de Sistemas, Pontificia Universidad Javeriana, Bogotá, Colombia
3
Doctorado en Ingeniería, Pontificia Universidad Javeriana, Bogotá, Colombia
Keywords: Decision Making, Adaptive System, Context-awareness, Dairy Ecosystem.
Abstract: Milk ecosystem of Boyacá department is an important economic sector; nevertheless, it presents economic
losses due to different problems, particularly, the lack of information to support decision making processes
regarding bovine diseases management. It must be emphasized that although there are solutions to this type
of problem, most of them only partially fulfill the required functionalities like: disease simulation systems in
particular regions, policy management, and notifications for different actors. For these reasons, we propose
EiBeLec, an adaptive system to support decision making, where users can visualize information according to
their requirements, context, characteristics and information needs, through services. In this position paper, we
describe EiBeLec and emphasize on the functionality provided to the government users (e.g., mayor,
governor, among others), supporting both decision making and regional infectious disease visualization, in
order to provide the authorities, the tools to generate policies and strategies for disease control and eradication.
1 INTRODUCTION
Computer technology provides tools for animal
production and reproduction, as well as for control
(Bradhurst et al., 2016; Ponge et al., 2016) and
eradication of infectious diseases that occur in
different livestock regions (Richards et al., 2014).
These tools use large data sets to generate reports that
serve as support for decision making by people
involved. However, the use of such data also requires
the development of policies, strategies and
approaches to select, analyze and interpret this data
appropriately.
An example of this issue can be found in the
region of Boyacá, Colombia. The dairy ecosystem of
Boyacá generates unstructured data represented in
different formats from different sources. For
example, (Mojica et al., 2007) reports a case in which
data are kept in spreadsheets, while in (Cruz et al.,
2014) data is taken from laboratory tests. A quick
search was done to find a tool capable of handling this
data, but, to the best of our knowledge, there is no
dedicated application for the dairy industry, that
integrates such data formats, and supports decision
making of final users.
In (Weeramanthri et al., 2010) planning is
integrated with decision making according to
government regulations. However, it does not take
into account the use of computer tools. On the other
hand, an economic health modeling tool focused on
cost savings linked to public health spending
priorities is presented in (Sanders et al., 2017);
nevertheless, it lacks results in practice, since the tool
is under development.
This highlights the need to create EiBeLec; an
adaptive information system to support decision
making where people involved in dairy ecosystem
can visualize information according to their
requirements, context, characteristics, and
information needs.
This paper focuses on the description of the
EiBeLec adaptive system architecture in terms of its
components and relationships. EiBeLec takes into
account different dairy ecosystem roles. However, for
the purposes of this paper, the role description is
limited to government roles (mayor, governor, among
others), in order to provide tailored services that help
in the decision making process regarding regional
disease evolution. To achieve this, EiBeLec considers
user profiles and its context. For example,
government actors should have a holistic view of the
Ballesteros-Ricaurte, J., Carrillo-Ramos, A., Acevedo, C. and Morantes, J.
Adaptive System to Support Decision-making of Dairy Ecosystem in Boyacá Department.
DOI: 10.5220/0006947502530260
In Proceedings of the 14th International Conference on Web Information Systems and Technologies (WEBIST 2018), pages 253-260
ISBN: 978-989-758-324-7
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
253

disease evolution in the whole region, while farmers
would only need the information for its vicinity.
This paper is organized as follows: section 2
presents works related to bovine diseases, support
systems for decision making and relationship
between data and tools used to present information.
Section 3 describes the EiBeLec adaptive model, each
layer and adaptation to content and display. In section
4, we present a case study emphasizing on the
government role. Finally, in section 5, we present the
conclusions and some pointers for future work.
2 RELATED WORKS
The analysis of related work is done from three
perspectives that involve areas such as bovine
epidemiology, adaptation, and use of information and
communication technologies (ICT).
Bovine viral diarrhea (BVD) is an endemic
disease in many countries; it leads to a variety of
health disorders that include mucosal and
reproductive problems (Martínez and Riveira, 2008,
Recamonde-Mendoza and Corbenilli, 2015).
Additionally, there is no complete information on
BVD prevalence in farms context (Cowley et al.,
2012) generating inconveniences for control and
eradication. The resulting impact on economic
productivity of cattle has turn BVD into a target for
control strategies in a selection of regions including
Austria, Scandinavia, Finland, Germany and
Switzerland (Tinsley et al., 2012).
BVD control and eradication programs in cattle in
the US have involved a collective decision-making
process on the best way to mitigate industrial losses.
Givens and Newcomer (2015) describe strategies
proposed by Livestock Industry that can be
implemented to control BVD considering diagnosis
for detection of infected fetus.
In (Santman-Berends et al., 2015) a model was
developed to predict BVD prevalence and subsequent
costs. The stability of the model outputs was
evaluated by comparing outputs of different numbers
of iterations. The model has an epidemiological
component in which a susceptible stochastic (S),
infectious (I), recovered (R) or vaccinated (V) model
was used to represent the BVD risk of incidence. The
output of the epidemiological component is used as
input for an economic component that uses
information about losses due to infections with BVD,
costs of vaccination, participation in BVD voluntary
eradication programs, and costs of different testing
and elimination of livestock. It is recommended for
BVD control scenarios to be implemented not only in
the dairy industry but also in the meat industry to
maximize the benefit of BVD control.
Table 1 shows works related to BVD and the use
of an epidemiological model to assess the disease's
contagion and strategies to control and eradication.
Table 1: Related works on infectious diseases.
Criteria
Machado et
al., 2015
Cowley et al.,
2012
Givens et
al., 2015
Santman
Berends et
al., 2015
Why use
epidemiological
model?
Prediction Analysis Analysis Prediction
What kind of
computer
application is it?
Statistical
program
Statistical
program
None Spread
sheet
The work defines
disease control
policies
Yes Yes Yes Yes
What are the
information
sources?
Surveys Private data Private
data
Private data
Context type that
takes into account
None Flock Farm Dairy farm
Who is the
application for?
N/A N/A
Livestock
Industry
Dairy farm
A dynamic model to specify socio-economic,
cultural, and ecological factors qualitatively is
proposed in (Mumba et al., 2017). It enables
individuals to identify spatial phenomenon directly
on physical maps, and integrate that information into
model development. These problems vary
considerably in space and context, suggesting that a
policy of a single size will not be effective. Therefore,
design and implementation of policies must consider
local needs in order to generate effective
interventions to control diseases within the
communities.
Armstrong and Kendall (2010) propose the
establishment of knowledge networks as a promising
method to support rapid adoption and generation of
health information regarding disease behaviour.
These networks will be particularly important for
implementation of national reform agenda,
responsive decision-making and translation of new
frameworks or competencies into practice. In
addition, it describes how interdisciplinary
knowledge networks establish a series of priority
areas of health research. Knowledge networks
composed of health professionals, decision makers of
health services, researchers, legislators and
consumers, who have the capacity to provide
approaches to build useful evidence at the point of
health service. After analysing the article, it can be
concluded that information technologies provide a
means by which knowledge can be stored and shared
by multiple users in different locations.
WEBIST 2018 - 14th International Conference on Web Information Systems and Technologies
254

Due to lack of updated information in livestock
census, decisions cannot be made on time and policy
planning is obstructed (Hollings et al., 2017). Auto-
learning methods have been proposed for estimating
livestock due to their potentially higher predictive
performance, and their ability to directly incorporate
complex interaction effects and noisy data (Robinson
et al., 2014; Elith et al., 2006). Advances in computer
skills, software and statistical innovation (Guisan and
Thuiller, 2005) have made machine learning
techniques provide support for decision making in an
emergency response situation.
Table 2 presents works related to relevance in
decision-making oriented to health issue, in this case,
public health (which takes into account human and
animal diseases).
Table 2: Projects related to decision making.
Criteria
Mumba et al.,
2017
Armstrong
and Kendall,
2010
Hollin
g
s et al.,
2017
Guisan and
Thuiller, 2005
There are
policies to
disease control
✓ ✓ ✓ ✓
Which role
makes
decisions?
Government Government Government Government
Who has the
responsibility to
plan strategies?
✓ ✓ ✓ ✕
Present
information
display
✕ ✓ ✓ ✕
It is important to mention that reviewed projects
indicate that there are two main problems: lack of
updated information and culture on the part of the
roles to report information. Taking into account that
these projects focus on relationship that can occur
between public policies, and control and eradication
of infectious diseases, there are no results that
articulate use of computer tools, infectious diseases in
cattle and policies of control and eradication; It can
also be evidenced, the lack of information about
context where infectious diseases are present. It is
clear that a system like EiBeLec is needed to fill this
gap. EiBeLec is an adaptive system that tests
visualization and information services to people
involved in dairy ecosystem of Boyacá in order to
support them in decision making. This system is
explained below.
3 EiBeLec
A phased methodology was followed to define
EiBeLec taking into account three models in a generic
way: the infectious disease, the dairy ecosystem, and
the context of use considering the adaptive model that
is explained below.
3.1 Definition of the Adaptative Model
The adaptive model has contextual variables of the
disease, the logical architecture of the adaptive
system, roles, and services offered by the system.
This model takes into account adaptation in terms of
content and deployment of information. Adaptation
model must consider data sources and information
updates. To achieve this, the model includes proces-
ses for disease context, dairy ecosystems, and roles.
Figure 1: Proposed architecture of the adaptive system.
Adaptive System to Support Decision-making of Dairy Ecosystem in Boyacá Department
255

Figure 1 shows the system architecture that is
composed of four layers. The lower layer corresponds
to Information sources, where adaptation data and
domain models are found. Contextual variables are
configured depending on infectious bovine disease
that is going to be simulated. In general, there are
variables common to diseases such as: temperature,
milking practices without good disinfection of
equipment, loan of breeders, artificial insemination,
high movement of livestock, among others. Some
information sources could be: government basic
sources (e.g. price regulations, farmer census,
topographical data, etc.), user profile repositories,
specific government sources (e.g. infection reports,
tests, treatment protocols, etc.). These data are
represented by means of two models: epidemiological
and contextual, which interpret data to provide
tailored services. Then, we find two complementary
layers: general and specific services, taking into
account the adaptation of information presented to
better suit the needs of the different user profiles.
Finally, there is the application layer to offer the
functionalities to the different roles that intervene in
dairy ecosystem.
3.1.1 Contextual Models
Context is related to: infectious disease, dairy
ecosystems, and role environment (see Figure 2).
However, for reasons of space, the latter will not be
detailed. Infectious diseases have specific symptoms,
which serve to determine what disease is occurring.
Variables that are related to symptoms are also related
to the context of the region where cattle are located.
In livestock farms there are several contextual pieces
of information that have to be taken into account like:
animals that mix with cattle, climate behavior,
geographical area where livestock farms are located.
Figure 2: Infectious disease context diagram.
For this paper, we chose the Bovine Viral
Diarrhea (BVD) as the disease that is presented in the
case study. BVD is a transmissible cattle disease,
characterized by fever, diarrhea, anorexia and cough,
which corresponds to an epidemic form of disease
like postnatal infection of BVD in susceptible herds,
also presenting intense diarrhea of short duration,
temporary decrease in the milk production and abor-
tions (Martínez and Riveira, 2008). Currently, BVD is
considered worldwide as one of the main diseases of
economic importance (Machado et al., 2013).
Transmission of BVD virus occurs in four ways:
horizontal, vertical, between herds, and within herd.
Figure 3 explains how horizontal transmission of
BVD virus occurs by: 1) direct contact with infected
animals, especially nose-nose, 2) contact through
saliva, semen, uterine secretions, placental fluids, and
3) contact with stool, urine, milk and waste.
Figure 3: Horizontal transmission of BVD.
Vertical transmission always occurs after
embryonic transmission, when a non-immune cow is
infected with BVD, a subclinical disease occurs and
the virus rapidly crosses the placenta (Martínez and
Riveira, 2008). Transmission between herds occurs
through acquisition of persistently infected cattle (PI)
or females that transport PI fetuses. Other routes of
introduction are: use of live vaccines, contaminated
semen, cohabitation with cattle, embryo transfer and
contact with bovines with acute infection. When an
infected animal is introduced to a herd, transmission
to susceptible animals occurs rapidly in most herd
animals; conversely, when infection is initiated with
a bovine with acute infection or by another route that
initiates an acute infection, transmission is of short
duration and only includes a small percentage of herd
before transmission ceases.
Figure 4: Dairy ecosystem context diagram.
WEBIST 2018 - 14th International Conference on Web Information Systems and Technologies
256

The dairy ecosystem of Boyacá is made up of
links such as cattle farms, producers, processors, and
milk transporters. Additionally, six roles that
intervene in the different processes have been
identified; the sum of these links and their
characteristics are taken into account in the context
description of the dairy ecosystem (see Figure 4).
Figure 5: Roles of dairy ecosystem of Boyacá.
Figure 5 shows the people involved in the dairy
ecosystem and how they are related both with
EiBeLec and with each other. People involved in
dairy ecosystem are very important to the system
since they fulfill two roles: feeding the system with
information about their activities, and acting as final
users of the system’s decision-support functionality.
The present paper focuses on the government role.
This role is illustrated in Figure 6 describing the
region of influence of government actors (i.e.
governors and mayors). The region is divided in
departments, municipalities, and towns. It is
important to notice that the mayor has influence over
the municipalities and their towns, while the governor
has influence over the whole region.
Figure 6: Relationship of government sector.
3.1.2 Services
Services are divided into specific and general, taking
into account presentation of information for roles,
configured from profiles.
The general services are of public consultation,
without the need of a specific profile; nevertheless,
specific services are provided according to the user
profile. Figure 7 shows an example where a
veterinarian uses general services, to verify a disease
outbreak, by means of consulting a disease in a
municipality and town. In the same way, the
Governor can make a specific query to see disease
behavior in the department.
Figure 7: Example of services.
3.2 User Profile of Government Role
Government sector corresponds to a group of actors
because it encloses the governor of the region, the
mayors of the municipalities that are part of dairy
ecosystem in the region, and the municipal and
departmental secretaries or bureau of health and
agriculture. This actor is the final user of the system,
since it requires information, reports, maps and a
description of information processed by the system.
The functionality of the system is twofold. On the one
hand, the government can use it to support decision
making processes, in order to create vaccination
campaigns, training for farmers in good practices on
livestock management, among others. On the other
hand, the government can also use the system to
detect and prevent possible epidemics, using
customized visualizations and alerts based on geo-
referenced data.
3.3 EiBeLec System
The proposed adaptive system is composed of
different components: information sources, context
data, models (epidemiological, social and adaptive),
services and actors, each of them has subdivisions and
activities that complement the system. In Figure 1 you
can see the structure and relationships between
Adaptive System to Support Decision-making of Dairy Ecosystem in Boyacá Department
257

components. Next, a diagram of the adaptive model
is described (see Figure 8).
Figure 8: Adaptive model diagram.
Information sources supply the system with data
and relationships through different artifacts; It is
divided into actors, exogenous sources and the context.
Context attributes must be considered because each
one obtains information from different scenarios, for
example, spatial-temporal that helps to obtain
information about physical geography of the region,
weather patterns, among others. The adaptive system
incorporates models that are directly related to sources
of information; they are responsible for processing
data, taking into account requirements of the system,
enabling stakeholders to access different queries
through general and specialized services so that they
can make decisions, from the role they fulfill.
4 CASE OF STUDY: EiBeLec
FOR GOVERNMENT
Figure 9: Visualization of virus in towns.
Suppose that, in one of the dairy municipalities of
Boyacá, there is a threat due to BVD outbreaks found
in some of their towns. In response, the EiBeLec
system should provide government actors with a
visualization similar to the map shown in Figure 9.
Cattle farms are found in different geographical areas,
they are close to rivers, lagoons, reservoirs of water,
they have different numbers of cattle in the paddocks
and might also have horses, goats, among others.
Several situations arise. Disease can spread
because farmers have no information about whether
their livestock are infected and consequently, they
could sell infected livestock in local markets. Also,
they might use the same places to feed the livestock.
They might as well take cattle to the same paddocks.
Even in the areas without virus alerts, farms might
be connected to infected cattle by one of the many
links like water reservoirs or milk transportation
routes that puts them at risk of contamination.
Farmers have unstructured information about
disease, causes and symptoms, economic losses
caused by disease in milk production, animal
reproduction, and purchase and sale of livestock; no
sanitary safety standards apply; and there is little
communication between farmers and government
sector. This results in a lack of updated information
in government officials for decision making, strategy
planning, and generation of policies of control and
eradication of this type of infectious diseases.
Government has different strategies to bring
information to farmers, but they have not been
effective (Ruiz, et al., 2012).
For these cases, three services are presented to the
government sector. The first service is to display
maps. Bovine infectious diseases occur in different
regions and propagation occurs using contextual
variables. This service presents information to
government sector using heat maps, taking into
account that the governor can see on the map how the
disease is distributed in the department, if there are
reservoirs of water near the focus of the disease,
which municipalities are affected, and which are
close to the disease. The second service consists of
notifying campaigns. Farmers do not have enough
information about control and eradication of bovine
infectious diseases. Authorities already organize
campaigns in different municipalities, but because of
the lack of interest or communication, farmers do not
take advantage of these campaigns. The early alert
service will notify government officials about BVD
outbreaks in their respective region of influence so
that they can act accordingly.
It also provides information about diffusion
campaigns aimed at farmers. The third service
corresponds to the generation of statistics. Decision
making is based on data that were processed. However,
the dairy sector does not have information systems to
capture, store, process and present reports. The service
of generating statistics provides information to the
authorities so that they can generate policies for control
and eradication of infectious diseases like strategies to
promote good livestock practices.
WEBIST 2018 - 14th International Conference on Web Information Systems and Technologies
258

To explain the use of services presented to the
government sector, two scenarios are described that
involve daily activities on livestock farms for
decision making by government sector.
Scenario 1: the governor wants to know if there is
an infectious disease that threatens public health; If
so, where is it happening and what is the behavior of
the disease considering: (1) how long ago it started,
(2) what can be its transmission estimation in
following days, (3) what aquifer systems are in the
area, and (4) what are the possible areas and causes
where the disease has started. To visualize it, one
must have information reported by different people
involved in the dairy ecosystem, which informs if
there are symptoms of diseases (see Table 3) using a
checklist, without the need to report the name of the
farm or its location.
Table 3: Check list of variables with disease.
Disease /
Variables
S T V X Y Z
BVD
     
Leukosis
     
Brucellosis
     
Fasciola
     
The list of variables that are related to the diseases
are: S: Abortions. T: Reuse of needles. V: Artificial
insemination. X: Use of non-certified semen. Y:
Mobilization of animals. Z: Improper handling of
farmyard.
Figure 10: Visualization for the Governor.
In this case, the Governor query the system,
clicking on the map (see Figure 10) and a heat map is
displayed, showing disease behaviour, zooming in
and viewing by regions.
Scenario 2: Vaccination campaigns, training and
data related to milk purchase prices are strategies that
serve to prevent, control and eradicate infectious
diseases. However, this information reaches a very
small group of farmers and there is no way to control
who is notified. In addition, messages do not reach all
farmers. Government sector has information to
generate campaigns to control and eradicate bovine
infectious diseases. Nevertheless, farmers do not have
access to information and communication with
government sector is not the most appropriate, since
one of the drawbacks is the lack of tools that allow
contacting farmers. Figure 11 shows the service of
message notification that government sector might
use to deliver information to farmers, to notify them
about different campaigns.
Figure 11: Display notification service.
5 CONCLUSIONS
This paper presents a bibliographic review of projects
related to bovine diseases, decision making and
people that intervene in dairy ecosystem. Decision
making by actors is a very important aspect in
processes of dairy sector. There must be a clear,
organized information and visualization of results and
arguments so that people involved can make better
decisions. However, to the best of our knowledge, the
systems that have been reviewed do not integrate
these services; therefore, we propose the construction
of EiBeLec, an adaptive system to support decision-
making, was proposed.
As a future work, the platform will be
implemented with the services proposed for the roles
of farmer and government. An example of service is
a simulation of an infectious disease spread in the
Boyacá department, taking into account the study of
automatic learning algorithms, performing tests to
determine which is the algorithm that presents the
best results with data of selected infectious diseases.
Furthermore, we propose an analysis of algorithms
(such as Logistic Regression (Pajares et al., 2010;
Chen et al., 2017; Richards et al., 2014), artificial
neural networks (Dwivedi, 2016), support vector
machines (Truong, Anh, Minh and Le, 2017)) for
processing of information, classification of symptoms
by means of different variables and the occurrence in
different livestock farms, data reported by actors.
This analysis will allow the selection of the most
appropriate algorithm to be used in EiBeLec.
ACKNOWLEDGMENTS
This research is financed by the Colombian
Government through the scholarship obtained in the
Adaptive System to Support Decision-making of Dairy Ecosystem in Boyacá Department
259

call number 733 of the department for science,
technology, and innovation (Colciencias) aimed to
the Formation of High Performance Human Capital.
REFERENCES
Armstrong, K., Kendall, E., 2010. Translating knowledge
into practice and policy: the role of knowledge
networks in primary health care. Health Information
Management Journal, 39(2): 9-17.
Bradhurst, R., Roche, S., East, I., Kwan, P., Garner, M.,
2016. Improving the computational efficiency of an
agent-based spatiotemporal model of livestock disease
spread and control. Environmental Modelling &
Software, 77: 1-12.
Cowley, D., Clegg, T., Doherty, M., More, S., 2012. Bovine
viral diarrhea virus seroprevalence and vaccination
usage in dairy and beef herds in the Republic of Ireland.
Irish Veterinary Journal, 65(1): 16.
Cruz, A., Moreno, G., González, K., Martínez, A., 2014.
Determinación de la presencia de anticuerpos contra
Neospora caninum y el virus de Dairrea Viral Bovina y
su relación con el desempeño reproductivo de hembras
bovinas del municipio de Oicatá (Boyacá). Rev CES
Med Zootec, 9(2): 238-247.
Chen, Y., Bi, K., Zhao, S., Ben-Arieh, D. y Wu, C., 2017.
Modeling individual fear factor with optimal control in
a disease-dynamic system. Chaos, Solitons & Fractals,
104: 531-545.
Dwivedi, A., 2016. Artificial neural network model for
effective cancer classification using microarray gene
expression data. Neural Comput & Applic, 29(12):
1545-1554.
Elith, J., Graham, C., Anderson, R., Dudik, M., Ferrier, S.,
Guisan, A., …Zimmermann, N., 2006. Novel methods
improve prediction of species distributions from
occurrence data. Ecography, 29(2): 129-151.
Givens, M., Newcomer, B., 2015. Perspective on BVDV
control programs. Animal Health Research Reviews,
16(1): 78-82.
Guisan, A., Thuiller, W., 2005. Predicting species
distribution: offering more than simple habitat models.
Ecology Letters, 8(9): 993-1009.
Hollings, T., Robinson, A., van Andel, M., Jewell, C.,
Burgman, M., 2017. Species distribution models: A
comparison of statistical approaches for livestock and
disease epidemics. PLoS ONE, 12(8): e0183626.
Machado, G., Egocheaga, R., Hein, H., Miranda, I., Neto,
W., Almeida, L., …Corbellini, G., 2013. Bovine Viral
Diarrhea (BVDV) in Dairy Cattle: A matched case-
control study. Transboundary and Emerging Diseases,
63(1): e1-13.
Machado, G., Recamonde-Mendoza, M., Corbellini, L.,
2015. What variables are important in predicting bovine
viral diarrhea virus? A random forest approach.
Veterinary Research, 24: 46-85.
Martínez, P., Riveira, I., 2008. Antecedentes, generalidades
y actualización en aspectos de patogénesis, diagnostico
y control de la Diarrea Viral Bovina (DVB) y
Rinotraqueitis Infecciosa Bovina (IBR). Trabajo de
grado, Pontificia Universidad Javeriana.
Mojica, F., Trujillo, R., Castellanos, D., Bernal, N., 2007.
Agenda prospectiva de investigación y desarrollo
tecnológico de la cadena láctea colombiana. Ministerio
de Agricultura y Desarrollo Rural.
Mumba, C., Skjerve, E., Rich, M., Rich, K., 2017.
Application of system dynamics and participatory
spatial group model building in animal health: A case
study of East Coast Fever interventions in Lundazi and
Monze districts of Zambia. PLoS ONE, 12(12):
e0189878.
Pajares, G., De la Cruz, J., 2010. Aprendizaje automático.
Un enfoque práctico. RA-MA Editorial.
Ponge, J., de Siquiera, D., Horstkemper, D., Hellingrath, B.,
Ludwig, S., Buarque, F., 2016. Automated scalable
modeling for population microsimulations. In
Conference on IEEE Symposium Series on
Computational Intelligence (SSCI 2016).
Robinson, T., Wint, G., Conchedda, G., van Boeckel, T.,
Ercoli, V., Palamara, E., …Gilbert, M., 2014. Mapping
the Global distribution of livestock. PLoS ONE, 9(5):
396084.
Richards, K., Hazelton, M., Stevenson, M., Lockhart, C.,
Pinto, J., Nguyen, L., 2014. Using exceedance
probabilities to detect anomalies in routinely recorded
animal health data, with particular reference to food-
and-mouth disease in Viet Nam. Spatial and Spatio-
temporal Epidemiology, 11: 125-133.
Ruiz, C., Henao, D., Lozano, M., Colorado, L., Mora, H.,
Velandia, J.,…Salazar, M., 2012. Plan estratégico
departamental de Ciencia, Tecnología e Innovación de
Boyacá. Observatorio Colombiano de Ciencia y
Tecnología – OcyT.
Santman-Berends, I., Mars, M., van Duijn, L., van Schaik,
G., 2015. Evaluation of the epidemiological and
economic consequences of control scenarios for bovine
viral diarrhea virus in dairy herds. Journal of Dairy
Science, 79(7): 1172-1181.
Sanders, T., Grove, A., Salway, S., Hampshaw, S., Goyder,
E., 2017. Incorporation of a health economic modelling
tool into public health commissioning: Evidence use in
a politicised context. Sociel Science & Medicine, 186:
122-129.
Tinsley, M., Lewis, F., Brülisauer, F., 2012. Network
modeling of BVD transmission. Veterinary Research,
43(1): 11.
Truong, P., Anh, N., Minh, N. y Le, L., 2017. A Machine
learning approach for drug discovery from herbal
medicine: Metabolite profiles to Therapeutic effects. In
Proceedings of the 8
th
International Conference on
Computational Systems-Biology and Bioinformatics
(CSBio 2017), 28-33.
Weeramanthri, T., Robertson, A., Dowse, G., Effler, P.,
Leclercq, M., Burtenshaw, J., …Gladstones, H., 2010.
Response to pandemic (H1N1) 2009 influenza in
Australia – lessons from a State health department
perspective. Australian Health Review, 34(4): 477-486.
WEBIST 2018 - 14th International Conference on Web Information Systems and Technologies
260