Analytics Applied to the Study of Reputational Risk through the

Analysis of Social Networks (Twitter) for the El Dorado Airport in

the City of Bogotá (Colombia)

Luis Gabriel Moreno Sandoval

and Liliana María Pantoja Rojas

Pontificia Universidad Javeriana, Facultad de Ingeniería, Colombia

Universidad Nacional de Colombia, Facultad de Ingeniería, Colombia

Keywords: Reputational Risk, Twitter, Polarity, Social Network, Computational Linguistic.

Abstract: Within a society increasingly technological and immersed in the different possibilities offered by the Internet

as a channel of interaction and communication, there is a need for private and state entities to explore the

information contained in social networks, an opportunity that implies new challenges associated with

emerging risks such as reputational. For this reason, through the study of social networks, reputational risk

and computational linguistics, an analysis of the strategic accounts of the El Dorado Airport of the city of

Bogotá (Colombia) on Twitter (@bog_eldorado) is carried out, taking into account the mentions and most

relevant hashtags, to identify the polarity through of BoW (Bag of Words) of your comments and infer the

user experience, customer satisfaction and how this influences reputational risk.

1 INTRODUCTION

The availability of massive amounts of data in the

scenario of the web has given new statistical and

computational impetus to the field of social network

analysis. According to Aggarwal (2011), thanks to

this emerging phenomenon, the study of social

networks has taken a new direction concerning the

results of data analytics, changes in the paradigms of

the computational social sciences along with the

analysis the text of social extraction, the complexity

sciences and social simulations.

Consequently, the computational challenges

associated with the ability to perform mining and

analytical processes to these sources of information in

the context of a social network, constitute an

unprecedented challenge and an opportunity to

determine useful information in a great variety of

fields (Mika , 2004; Mislove, Marcon, Gummadi,

Druschel, and Bhattacharjee, 2007; Zafar,

Bhattacharya, Ganguly, Gummadi, and Ghosh, 2015;

Zafarani, Abbasi, and Liu, 2014), including

reputational risk.

Participation in contemporary social life through

networks has become increasingly complicated;

people accumulate hundreds of friends and

acquaintances through social media, but more and

more report a smaller number of friends they can trust

(Burt, Kilduff, and Tasselli, 2013).

For the foregoing and in order to analyze as a case

study the reputational risk at El Dorado Airport in

Bogotá (Colombia) taking into account the most

relevant mentions and hashtags, to identify the

polarity of their comments and infer the user

experience and customer satisfaction, six sections are

presented. The second refers to the conceptual

framework, the third is the case study. In the fourth

and fifth sections is the methodology and

development and in the sixth section the conclusions

and future work.

2 CONCEPTUAL REVIEW

The present study has three fundamental pillars:

reputation, social networks and computational

linguistics that will be described below:

2.1 Reputational Risk

The traditional types of risk that can affect reputation

are mainly strategic, operational, financial and

compliance risks, which can often be quantified, but

their collective impact on reputation is difficult to

488

Moreno Sandoval, L. and Pantoja Rojas, L.

Analytics Applied to the Study of Reputational Risk through the Analysis of Social Networks (Twitter) for the El Dorado Airport in the City of Bogotá (Colombia).

DOI: 10.5220/0007770804880495

In Proceedings of the 21st International Conference on Enterprise Information Systems (ICEIS 2019), pages 488-495

ISBN: 978-989-758-372-8

foresee (Lamont, 2015). However, reputation can also

be affected by additional risks that fall outside these

areas, especially those that come from social

networks.

It should be understood that reputation is one of the

main intangible assets that helps organizations not

only to obtain competitive advantages, but also to

survive in times of economic turbulence (Fernández-

Gámez, Gil-Corral, and Galán-Valdivieso, 2016),

which is why it is necessary for organizations to use

a multifaceted approach.

The organization managers, estimate that an

intangible increasingly important in the

organizational field is the corporate reputation (CR),

and although it is the most important asset, it is also

the most difficult to protect (E., De la Fuente Sabaté,

and Delgado García, 2005). In the eighties, many

experts in organizational valuation assured that 70%

of the value of a business depended on its tangible

assets and 30% of what its intangible assets

contributed. At present, it is thought that 75% of the

value of a business depends on the quality and

management of intangible assets (Tomás Garicano

Rojas, 2011).

Now, for Hernández Velasco (2012), from the

corporate point of view, reputation is the recognition

that interest groups or "stakeholders" make of the

behavior of an organization in the satisfaction of its

expectations and reputational risk in relation to the

response that these "stakeholders" can have when

their expectations are not met.

2.2 Social Network Analysis (SNA)

A social network consists of a set of relationships

established between a group of social actors, as well

as any additional information about those actors and

their relationships (Ö Bodin and Prell, 2011). The

social network's approach is based on the notion of

having patterns associated with social ties, in which

some actors are involved, and their interactions have

significant consequences for them. Social network

analysts seek to discover different kinds of patterns,

determine the conditions under which they arise and

explore their consequences on the attitudes,

perspectives, and behaviors of individuals, groups or

subgroups and the systems to which they belong

(Scott J, 2011 ).

2.1.1 Basic elements of a Network

According to Bodin and Crona, (2009), we have the

following definitions of the essential elements of a

network:

Nodes or actors: These are the people or groups of

people who are around a common goal. Usually, the

nodes or actors are represented by circles. The sums

of all the nodes represent the size of the network.

Link: These are the links that exist between two or

more nodes. In a network, a direct connection with

another actor is shown. Lines represent the links or

relationships.

Flow: Indicates the link address. An arrow indicating

the direction represents the streams. It is also possible

that there are also mutual or bidirectional flows.

When an actor does not have any current, which in

turn implies no link, it is said that this node is loose

within the network.

2.1.2 Analysis of Social Networks-Properties

According to Freeman (2004), the analysis of social

networks is based on four characteristics that have

been consolidated since their appearance and whose

integration gives rise to a research paradigm:

a) The study of social networks is motivated by a

basic intuition based on the bonds that unite

social actors.

b) It is an approach that is based on systematic and

empirical data.

c) It is found in no small extent on graphic

representations and,

d) It relies on the use of mathematical and

computational models.

According to the thesis: "Evolution of the policy

network for innovation in Colombia: The emerging

phenomenon of confirmation of governance

networks" (Moreno, 2015), the structural

characteristics of the network that the theory on the

analysis of social networks distinguish the following

elements:

Level of cohesion of the network: an essential feature

of social networks is their level of coherence, that is,

to what extent the network "works together" when

presenting relations between nodes instead of being

divided into subgroups cohesive among themselves.

This means that a network with high structural

cohesion lacks a set of differentiable subsets,

understanding that the existence of subgroups

represents challenges to unite the action towards

common initiatives.

Position and influence of the network: this is a

structural characteristic that is analyzed to understand

the location of the actors in the structure to influence

the system. Being better placed can access valuable

information (Burt et al., 2013). There are several

ways to define and measure centrality in social

Analytics Applied to the Study of Reputational Risk through the Analysis of Social Networks (Twitter) for the El Dorado Airport in the City

of Bogotá (Colombia)

489

networks. The first is by a degree of centrality and the

second by a degree of intermediation.

The degree of centrality refers to the number of links

an actor has or can be understood as the degree to

which an individual actor connects with other actors

and is associated with the level of activity in the

network (Burt et al., 2013). In the analysis of social

networks, the degree to which an actor indirectly

connects to other actors is often quantified using the

degree of intermediation (Freeman, 2004) with which

it is determined that an actor that is among many other

actors in The network has a high degree of mediation,

which implies that it could act as a bridge between

others that would be disconnected without their

existence.

2.3 Computational Linguistics

According to Montes-y-Gómez, (2002)

computational linguistics is "the science that deals

with the application of computational methods in the

study of natural language. This science is a

combination of two larger sciences; linguistics,

which studies the laws of human language, and

artificial intelligence, which investigates

computational methods for the management of

complex systems. "

Now, the sentimental analysis aims to determine the

attitude of a speaker or a writer concerning some

topic, or the global polarity of a document. According

to Albornoz (2011) the polarity classification aims to

obtain a score that indicates whether the text

expresses a positive or negative opinion, within a

range where 0 would mean a neutral subjective load,

1 a positive personal load and -1 a load personal

negative.

It is possible to use many techniques in the

classification of polarity, these can be supervised

algorithms such as SVM (Support Vector Machine),

LR (Logistic Regression), NB (Naive Bayes) among

others, there are other types of models are based on

lexicons; in this case study was used a polarity

lexicon called "CSL: A Combined Spanish Lexicon -

Resource for Polarity" through a process of BoW

(Bag of Words) (Moreno et al., 2017). It was

possible to determine the polarity of comments,

hashtags, and mentions in the tweets, finally, these

results are consolidated in the kind cluster file that is

read for the SNA process.

3 CASE STUDY: EL DORADO

AIRPORT

The El Dorado Airport is part of the urban and

architectural history of Colombia and Bogotá.

OPAIN S.A. is the company set up with the sole

objective of managing, modernizing, commercially

developing, expanding, operating and maintaining

the El Dorado International Airport.

It is the first airport in Latin America in cargo volume

and the third largest airport in Latin America in

passenger volume, after the International Airport of

Mexico City and the International Airport of São

Paulo-Guarulhos.

He won the Skytrax award for the best personnel in

South America in 2016, as well as the first place on

the list of the best airports in Latin America for the

second time in a row. On the other hand, in the list of

the best airports in the world, El Dorado achieved the

42nd position in the list of 2017.

A total of 42 lines arrive at the airport, discriminated

against as follows: 21 international airlines, six

domestic airlines, and 15 cargo airlines.

4 METHODOLOGY

This is empirical research, that is, based on

experimentation, to test the hypothesis that under the

evolution of the structural properties of the social

network Twitter it is possible to establish

relationships that show reputational risk.

The analysis of the structural properties of the El

Dorado Airport network (as a case study) is done by

systematizing and examining the relationships

established through Twitter, for which the

methodological design is based on the use of the

Pajek and Voswiever software for the construction of

graphs corresponding to each of the structural

properties defined for the study. All this, from the

data obtained by "crawling" (crawling) the public

access information of Twitter through the use of

application programming interfaces (APIs) whose

complexities describe Mislove et al. (2007)

extensively.

This is how to achieve the objectives of this empirical

research is carried out in three phases in the

methodological design:

a) Conceptual exploration for the identification of

reputational risk categories and structural

analysis.

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

490

b) Application of an exploratory methodology to

describe the structural properties of the El

Dorado Airport's digital ecosystem.

c) Empirical evaluation of the behavior of the

network in its general interactions of the El

Dorado Airport network.

The steps to carry out the following investigation are

summarized below:

1. A network is woven to analyze it empirically,

and the social structure is explained to establish

its degree of significance and relevance for the

case study.

2. The types of account to be analyzed are found

because they are not homogeneous since they are

mixed reports of organizations with personal

accounts.

3. All friendship links are "downloaded" on Twitter

and friends are established for each of the

accounts.

4. The "mentions network" and the "hashtag

network" are established.

5. The user is chosen, and the tweets are taken, and

then each of the posts is decomposed, and the

polarity is calculated (mentions and hashtag).

6. The list of remarks and hashtag is stored; the

value of the vector is normalized. The amount of

mentions and hashtag of the moment is

accumulated in a vector per user mentioned, and

the number suggested to determine the

importance.

The normalized value is equal to:

normalized value = (ln value to normalize + 1)

/ (ln (calculated max - min calculated + 1)) *

50)

7. Each post is read and polarity is extracted

through the iSOL and CSL Caoba lexicon,

accumulating the polarity of the user. The

information is accumulated as positive, negative

and neutral.

8. Pajek files are written.

9. VOSviewer is used; According to its official

website http://www.vosviewer.com/, as a

software tool for the construction and

visualization of networks.

5 DEVELOPING

The present study was conducted during April and

May of 2018. The ecosystem of social network

accounts: 72 (Appendix)

The data collected are the following:

Table 1: Dates.

Description

Count

Follower

41826

Hashtag

65013

Mention

146922

Profile

41841

Tweets

176251

Phase 1: The study is closed for Twitter, all tweets

from the ecosystem timeline of social media accounts

were taken, and followers, hashtags and mentions at

the structural level were analyzed, showing the

density of the interconnections between the followers

of the reports of the digital ecosystem of El Dorado

Airport.

Phase 2 Polarity is defined as positive, negative and

neutral.

According to the information analyzed, the following

graphs are obtained.

Figure 1: Followers.

The graph shows how the density of the

interconnections between the followers of the digital

ecosystem accounts of the El Dorado Airport has

manifested and also suggests that the links of the

topics discussed are stable due to a large amount of

message exchang.

Analytics Applied to the Study of Reputational Risk through the Analysis of Social Networks (Twitter) for the El Dorado Airport in the City

of Bogotá (Colombia)

491

Figure 2: Degree- El Dorado Airport.

On Twitter, the nodes with the most significant

number of followers when issuing a trill, have a

higher probability of generating an impact on the

network through the simple fact of reaching a more

substantial amount of users. This impact does not

depend solely on the number of followers but the

number of responses or mentions generated as a result

of the trills of the user.

Therefore, it is evident that Air Canada, American

Air, AirFrance, Aviatur and Avianca are the ones

with the highest degree of centrality concerning the

airport in question.

Figure 3: Between Centrality - El Dorado Airport.

Nodes with high intermediation play an essential role

in the communication of the entire network.

Therefore, it is observed that the nodes that connect

more effectively within the system are again Air

Canada, American Air, AirFrance, Aviatur and

Avianca.

Figure 4: Clonnes Centrality – El Dorado Airport.

The centrality by proximity determines how close a

node is to all the other nodes of the network,

establishing the importance of calculating the

distance that exists between the node with the rest of

the system and a node with greater proximity. The

graph shows great importance with Avianca and

AirFrance.

Hashtags are essential for any digital marketing

campaign that is carried out, as they can frame the

success or failure of an idea, slogan or concept of any

brand, product or organization. Outstanding

#AttendeesTravellers, #Happy Monday, #Cofee

Figure 5: Polarity Hashtag - El Dorado Airport.

When observing the polarity of "AtentosViajeros", it

is found that it is positive because it allows the users

to know the situations of the last moment of the

airport and to be attentive to the events that are

presented. When analyzing

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

492

#AttendeesTraveling Our Warriors of the

@FCFSeleccionCol are coming home and we are

ready to receive them. If you travel from

@BOG_ELDORADO tomorrow, keep in mind that

congestions and vials closings are planned on 26th

street from 10:00 a.m.

But for #HappyMonday it is observed that the

algorithms used are able to abstract the sarcasm since

the message is opposed to the hashtag

#HappyMonday #Holiday less for airport users the

golden one who are locked in a plane waiting for a

runway #Planretorno

The coffee for example that indicates. "I just copied

an aromatic and a cake in the airport Oma and then I

was left without the rent"

The foregoing demonstrates aspects that the Airport

did not contemplate but that its partners were

affecting their reputation.

Figure 6: Polarity of Mentions.

When carrying out the polarity analysis, for this case

a reputational risk contagion issue is observed

between the relationship of Aerocivil de Colombia

and El Dorado Airport; this was due to the fact that

situations such as the lack of light at the airport were

attributed to the airport, although it is an obligation of

Aerocivil; however, users directly related the

responsibility to El Dorado because there is no

distinction between the public of what activities each

of these entities performs.

6 CONCLUSIONS

Social networks have great potential to identify

groups of interest, know their needs, expectations,

tastes, preferences, and opinions on specific

initiatives and / or products; therefore, this type of

studies allows identifying events that can generate

reputational risk in an organization.

Currently, the volume of comments on the network is

very high, so an organization does not have the

capacity to process them manually; therefore, it is

necessary to have technological tools such as those

presented in this study, which help to automatically

transform all the opinions of interest and identify their

type of polarity (positive, negative or neutral) for

decision making.

It is possible the integration of computational

linguistic analysis with the structural analysis of

networks linking the results of the polarities

associated with the mentions and hashtag, which

allows understanding the dynamics of the reference

risk information. To the extent that actions to control

reputational risk are identified, monitored and

executed, the gap closes, and the probability that the

crisis in an organization is managed will be much

more effective.

The information contained in social networks is an

opportunity that involves new risks associated with

risks such as reputational risk. Through the SNA on

Twitter and computational linguistics, it is possible to

identify the polarity of comments and hashtag to infer

the user experience and customer satisfaction.

According to the structural analysis, foreign airlines

such as AirCanada and AirFrance are fundamental for

strategic alliances in the handling of information

since the reputation comes from third parties and

what is sought with the protection of the status is the

creation of value of an organization.

The polarity of the comments and hashtag as well as

the identification of topics at El Dorado Airport

allowed us to find aspects that the Airport was not

contemplating but that were affecting its reputation as

Analytics Applied to the Study of Reputational Risk through the Analysis of Social Networks (Twitter) for the El Dorado Airport in the City

of Bogotá (Colombia)

493

the coffee issue related mainly to prices and service,

which was changing his status.

In conclusion, this work contributes in the emerging

study of reputational risk through the Twitter network

is a very nascent topic in Latin America; therefore, it

should be considered as a new field of research to be

used by different organizations, since the reputation

is of the main assets intangible of a company.

For future work, the reputation risk can be supported

by supervised techniques of polarity classification,

this would help to improve the calculation of risk in

digital social networks. It has structural information

which can be integrated into the training data set as

the number of followers or the average number of

words and sentences for tweets, adding more

characteristics in the models of polarity classification.

ACKNOWLEDGEMENTS

This research was carried out by the Center of

Excellence and Appropriation in Big Data and

Data Analytics (CAOBA). It is being led by the

Pontificia Universidad Javeriana Colombia and it

was funded by the Ministry of Information

Technologies and Telecommunications of the

Republic of Colombia (MinTIC) through the

Colombian Administrative Department of Science,

Technology and Innovation (COLCIENCIAS)

within contract No. FP44842-anex46-2015.

REFERENCES

Aggarwal C.C. (2011) An Introduction to Social Network

Data Analytics. In: Aggarwal C. (eds) Social Network

Data Analytics. Springer, Boston, MA.

Becerra Bonza, D. C., and Guerrero García, T. M. (2012).

Social representations workers of a public organization

have about psychosocial risks and harassment in the

workplace towards the Promotion of Health, 17 (1),

118-132.

Bodin, Ö., and Crona, B. I. (2009). The role of social

networks in natural resource governance: What

relational patterns make a difference ?. Global

environmental change, 19 (3), 366-374.

Bodin, Ö., and Prell, C. (Eds.). (2011). Social networks and

natural resource management: uncovering the social

fabric of environmental governance. Cambridge

University Press.

Bourcier, D. (2003). Artificial intelligence and law.

Editorial UOC

Burt, R. S., Kilduff, M., and Tasselli, S. (2013). Social

Network Analysis: Foundations and Frontiers on

Advantage. http://doi.org/10.1146/annurev-psych-

113011-143828.

Contreras Barrera, M. (2014). Text mining: a current vision

University Library, vol. 17, no. 2, July-December,

2014, pp. 129-138 National Autonomous University of

Mexico Federal District, Mexico. University Library,

17 (2), 129-138

From Albornoz, J.C., Plaza, L., Gervás, P., and Díaz, A.

(2011, April). A joint model of feature mining and

sentiment analysis for product review rating. In

European conference on retrieval information (pp. 55-

66). Springer, Berlin, Heidelberg

E., from the Sabaté Source, J. M., and Garcia, J. B. D.

(2005). Corporate reputation and value creation. A

theoretical framework of a circular relationship.

European Investigations of Management and Economy

of the Company, 11 (2), 81-97.

Fernández-Gámez, M.A., Gil-Corral, A.M., and Galán-

Valdivieso, F. (2016). Corporate reputation and market

value: Evidence with generalized regression neural

networks. Expert Systems with Applications.

http://doi.org/10.1016/j.eswa.2015.10.028

Freeman, L. (2004). The development of social network

analysis. A Study in the Sociology of Science, 1.

Garicano, T. (2011). Corporate governance and corporate

reputation. Corporate Governance Center

http://doi.org/10.1016/j.dss.2013.09.004Hernández

Velasco, J. (2012). Reputational risk: Assess to prevent;

prevent to act. Give in, 32, 1-10.

Hernández M, and Gómez, J. (2013). Natural Language

Processing Applications. Polytechnic Journal, Vol. 32,

N, 87-96.

Kanayama, H., and Nasukawa, T. (2006, July). Fully

automatic lexicon expansion for domain-oriented

sentiment analysis. In Proceedings of the 2006

conference on empirical methods in natural language

processing (pp. 355-363). Association for

Computational Linguistics.

Kim, S. M., and Hovy, E. (2004, August). Determining the

sentiment of opinions. In Proceedings of the 20th

international conference on Computational Linguistics

(page 1367). Association for Computational

Linguistics.

Lamont, J. (2015). Risk Management: Reputation is

everything. KMWorld, 60 (1), 8. Retrieved from

http://ra.ocls.ca/ra/login.aspx?inst=conestoga&url=htt

p://search.ebscohost.com/login.aspx?direct=true&db=

bth&AN = 100327144 & site = bsi-live & scope = site.

Martín De Castro, G. (2008). Business reputation and

competitive advantage. Esic Editorial. Madrid. pp. 32-35.

Mika, P. (2004). Social networks and the semantic web.

Proceedings of the 2004 IEEE / WIC / ACM

International Conference on Web Intelligence.

http://doi.org/10.1109/WI.2004.10039

Mislove, A., Campus, E., Marcon, M., Gummadi, K. P., and

Druschel, P. (2007). Measurement and Analysis of

Online Social Networks.

Montes-y-Gómez, M., Gelbukh, A., and López-López, A.

(2002, July). Text mining at detail level using

conceptual graphs. In International Conference on

Conceptual Structures (pp. 122-136). Springer, Berlin,

Heidelberg.

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

494

Moreno-Sandoval L. (2015). Master's Thesis: Evolution of

the policy network for innovation in Colombia: The

phenomenon emergent form of governance networks.

Faculty of Administration. university externship of

Colombia.

Moreno-Sandoval L., Beltrán-Herrera P., Vargas-Cruz J.,

Sánchez-Barriga C., Pomares-Quimbaya A., Alvarado-

Valencia J. and García-Díaz J. (2017). CSL: A

Combined Spanish Lexicon - Resource for Polarity

Classification and Sentiment Analysis.In Proceedings

of the 19th International Conference on Enterprise

Information Systems - Volume 1: ICEIS, ISBN 978-

989-758-247-9, pages 288-295. DOI:

10.5220/0006336402880295

Pang, B., Lee, L., and Vaithyanathan, S. (2002, July).

Thumbs up ?: feeling classification using machine

learning techniques. In Proceedings of the ACL-02

conference on Empirical Methods in Natural Language

Processing, Volume 10 (pages 79-86). Association for

computational linguistics.

Scott, J. (2011). Social network analysis: developments,

advances, and prospects. Social Network Analysis and

Mining, 1 (1), 21-26. http://doi.org/10.1007/s13278-

010-0012-6

Tomás Garicano Rojas. (2011). Introduction to the

management of reputational risks. IE Bussines School

and Corporate Reputation Forum, 1-97.

Zafar, M. B., Bhattacharya, P., Ganguly, N., Gummadi, K.

P., and Ghosh, S. (2015). Sampling content from online

social networks: Comparing random vs. expert

sampling of the twitter stream. ACM Transactions on

the Web (TWEB), 9 (3), 12.

Zafarani, R., Abbasi, M.A., and Liu, H. (2014). Social

media mining: an introduction. Cambridge University

Press

APPENDIX

Analyzed Digital Ecosystem Airport El Dorado

bog_eldorado

xuecafe

iemark

vivaaircolombia

aerocambiar

cambiosdorado

interjet

wingstop

aerolineas_ar

carolinaherrera

jetblue

ziropevzla

aeromexico

colsubsidio_ofi

klm

cancilleriacol

aviatur

compassgroupcol

kokoriko_col

migracioncol

aircanada

prioritypasscom

latam_co

diancolombia

aireuropaco

copaairlines

lasamigration

icacolombia

airfrance

crepeswafflesco

lufthansa

policiabogota

americanair

davivienda

unibaggage

policiaaduanera

assistcard

directvâ

mh_colombia

policiaantinar

attenzadf

worlddutyfreees

mcdonalds

aerocivilcol

avantel_sas

dunkindonuts

menziesaviation

redirecciona el

dorado

avianca

easyflyvuelos

omacolombia

invimacolombia.

beumergroup

elcorral_

aerolineasatena

trafficaircolom

cafebritt

elmarketco

spiritairlines

anatonacional

budget

frisbylohace

starbucks

reportcolombia

burgerking

ussfontibon

tame_ep

idubogota

cafeygourmet

iberia

turkishairlines

policiacolombia

Analytics Applied to the Study of Reputational Risk through the Analysis of Social Networks (Twitter) for the El Dorado Airport in the City

of Bogotá (Colombia)

495