=Paper=
{{Paper
|id=Vol-2346/paper2
|storemode=property
|title=Towards an Argumentation System for Assisting Users with Privacy Management in Online Social Networks
|pdfUrl=https://ceur-ws.org/Vol-2346/paper2.pdf
|volume=Vol-2346
|authors=Ramon Ruiz-Dolz,Stella Heras,Jose Alemany,Ana Garcia-Fornes
|dblpUrl=https://dblp.org/rec/conf/persuasive/Ruiz-DolzHAG19
}}
==Towards an Argumentation System for Assisting Users with Privacy Management in Online Social Networks==
https://ceur-ws.org/Vol-2346/paper2.pdf
Towards an Argumentation System for Assisting
Users with Privacy Management in Online
Social Networks
Ramon Ruiz-Dolz[0000−0002−3059−8520] , Stella Heras[0000−0001−6212−9377] , Jose
Alemany, and Ana Garcia-Fornes[0000−0003−4482−8793]
Departament de Sistemes Informàtics y Computació
Universitat Politècnica de València
Camino de Vera s/n. 46022 Valencia (Spain)
{raruidol, sheras, jalemany1, agarcia}@dsic.upv.es
Abstract. In this paper, we present an argumentation system for sup-
porting users of online social networks on their decision making. The
purpose of the argumentation system is to give a reasoned recommenda-
tion to a specific user on whether he/she should or should not perform
an action in the social network (e.g. post a comment, share a photo) tak-
ing into account his/her privacy preferences. We both define the argu-
mentation framework and the architecture of the argumentation system.
Finally, we illustrate the system with a practical example in a pedagogic
social network.
Keywords: Argumentation · Persuasion · Social Networks · Privacy ·
Security
1 Introduction
One of the most concerning dangers of the last years is the unconscious sharing
of sensitive information in online social networks (OSN). The main purpose of
this work is to define the basis of an argumentation system that detects when
sensitive data is being published, and persuades users to modify the content
of the publication. Two main types of privacy threats can happen in an OSN.
An indirect threat may occur when a user makes a publication involving other
users in the post. It is possible that this publication violates the privacy policy
of any of its members. On the other hand, a direct threat happens when the
author of the publication violates its own privacy preferences. Even though this
may seem incoherent, it is a usual threat in OSNs, since users commonly inter-
act disregarding what they share or with who they share sensitive information.
Our argumentation system can deal with both types of threat, and can help to
minimise the number of undesired publications of sensitive data.
This paper is structured as follows, in section 2 we briefly survey related
work; in section 3 we define our argumentation system; in section 4, we propose
an example of use of the system; finally, in section 5 we summarise the conclusions
and our future lines of work.
2 R. Ruiz-Dolz et al.
2 Related work
Arguing is the natural way for humans to explain themselves to resolve disputes
and reach agreements. Thus, computational argumentation is a powerful tool
that allows intelligent systems that interact with humans to generate arguments
to explain their behaviour and hence, persuade their users to accept it (e.g.
decisions, recommendations). In the specific domain of OSN, computational ar-
gumentation can help to structure user opinions and enhance dialogues [9], to
model the dialogue between users that share their recommendations in the net-
work [10] or to prevent users from unconsciously compromise their privacy or
that of others [18].
By being able to generate persuasive arguments and engage with the user in
a persuasion dialogue, the OSN can warn users of the implications and conse-
quences of their actions. Therefore, we can find in the literature recent research
that follows this trend and applies computational argumentation techniques in
the context of preserving the users’ privacy in this type of networks. For in-
stance, in [11] authors propose an assumption-based argumentation model that
allows agents representing users in a social network to argue and decide whether
a content should be shared or not. To generate arguments, agents make use
of semantic rules that represent their users’ privacy constraints. In this work,
authors assume that an agent is aware of the semantic rules of its user. How-
ever, our system is able to automatically generate arguments by using the users’
information gathered from the social network.
A similar work [7] applied the theory of argumentation schemes to negotiate
with users the setting of a privacy preference, thus managing multiparty privacy
conflicts based on logical arguments. Users employ arguments to convince the
other parties that their demands are reasonable and should be taken into ac-
count. However, this work is centred on presenting the theoretical model, but it
does not specify how arguments are actually generated and evaluated. In a sub-
sequent work [6], authors proposed a novel model for representing and reasoning
about multiparty privacy conflicts by using contextual factors, user preferences,
and user arguments. Here, argumentation is conceived as in internal process by
which agents can resolve privacy disputes once this problem occurs, but not as
a persuasion tool that the system may use to interact with the user to help
him/her to make good decisions and avoid privacy conflicts.
3 Argumentation system
In this section, we define our argumentation system as an educational tool to
preserve the privacy of the users of an OSN. To this end, we assume that the
system operates in a OSN that includes the common features of this type of
networks (e.g. user information and preferences, friends, groups, privacy config-
uration) and that allows users to perform common social actions (e.g. posting a
comment, sharing a photo) [16].
An Argumentation System for Assisting Users in Privacy Management 3
3.1 Framework formalisation
Our Argumentation Framework is based on Quantitative Bipolar Argumentation
Frameworks [2].
Definition 1 (Argumentation Framework for Online Social Networks).
We define an argumentation framework for online social networks as a tuple
AF OSN = where: A is a set of n arguments [α0 , . . . , αn ]; R is
the attack relation on A such as A×A → R; P is the list of e profiles involved in
an argumentation process [p0 , . . . , pe ]; and τp is a function A × P → [0, . . . , 1]
that determines the score of an argument α for a given profile p.
In our framework, each individual argument α = (β, T , D) is defined by
three parameters. β is the claim or bias of the argument. It is represented as a
binary variable that determines whether an argument acts in favour or against
performing an action in the social network. T is the label of the argument, which
represents the four different types of arguments that can be generated by our
argumentation system: Privacy, Trust, Risk and Content arguments. Finally, D
is the support of the argument. This parameter consists of a value derived from
all the information gathered from the social network in order to infer the claim
of a determined type of argument, and hence depends on the type of argument.
Each relationship r = (αi , αj ) represents an attack from αi towards αj . As
proposed in [12] and [13], a rebuttal attack occurs when an argument invalidates
other argument’s claim (e.g. α1 = (-1, T1 , D1 ) rebuts α2 = (+1, T2 , D2 ) and vice
versa). On the other hand, an undercut attack is carried out when an argument’s
claim invalidates other argument’s support. The internal argumentation process
performed by our system to generate the acceptable arguments for a particular
conflict in the OSN only allows rebuttal attacks. In the human interactive argu-
mentation process carried out by our dialogue module both types of attacks will
be considered.
Regarding the user profile, we define p = (ν,ρ,M ) as the combination of the
preference values ν, the personality ρ and a list of miscellaneous information M .
The preference values ν is a vector containing the preferences that each user
has towards a concrete value that the arguments of our system may promote
[3]. We propose the following preferences based on [15] to be considered in our
system: Privacy/Popularity, Closeness/Openness, Flexibility/Intransigence and
Content Sensitivity. The first three bipolar preferences P /P , are defined as a
value v in the [0,1] range, being v the value assigned to P and (1 - v) to P .
Therefore, a user profile with Privacy/Popularity = 0.2 would have a 0.2 prefer-
ence for Privacy and (1 - 0.2) preference for Popularity. The Content Sensitivity
preference is defined as a value v in the range [0, 1] being 1 the maximum con-
cern about the content sensitivity and 0 if the user does not really care about
this preference. This value is calculated as the average of the 6 different types of
content considered in this work.
The personality of a user profile ρ is a 5 dimension vector that models the
personality of a specific user based on the five parameters proposed in [14]. The
personality dimensions taken into account are the Openness, Conscientiousness,
4 R. Ruiz-Dolz et al.
Extraversion, Agreeableness and Neuroticism. Here we assume that this infor-
mation is available since users of the network have undertaken a personality test
or else, that these dimensions can be automatically determined by the activities
of the user in the social network [8].
The last part of the user definition (M ) is a set of general information ex-
tracted from a specific user profile such as the age, location, likes, etc.
Finally, we define the scoring function τp as the function that takes an ar-
gument and a profile as input and determines the value of the argument in the
context of a specific user profile. In order to obtain this score, function τp is
defined as,
τp (α, p) = αβ · αD · pν (1)
Thus, the score of an argument for a specific user profile is basically the
product of the support value of the argument, the preference value that promotes
the argument and the bias of that argument.
Definition 2 (Defeat). An argument αi ∈ A defeats another argument αj ∈ A
in a context determined by a user profile p iff (αi , αj ) ∈ R ∧ |τp (αi , p)| > |τp (αj ,
p)|.
Then, we can define def eatp (αi , αj ) if there exists an attack relationship
between both arguments and the score of the argument αi is higher than the
score of the argument αj . It means that the argument αi is promoting a value
that is preferred by the user that receives the argument.
Definition 3 (Acceptability). An argument αi ∈ A is acceptable in a context
determined by a user profile p iff ∀ αj ∈ A ∧ def eatp (αj , αi ) → ∃αk ∈ A ∧
def eatp (αk , αj ).
In other words, an argument is acceptable if there are no other undefeated
arguments attacking it.
3.2 System architecture
An argumentative process is defined by the achievement of four main tasks:
identification, analysis, evaluation and invention [17]. Our system, graphically
depicted in Figure 1 consists of four different modules designed to perform all
these essential tasks.
Feature extraction module. This module is in charge of obtaining all the
relevant information for our framework directly from the OSN. The information
obtained by this module will be used to model user profiles and to get the
parameters that will define an argument in our system.
The main purpose of modelling user profiles is to be able to learn which
arguments are more persuasive to what type of user. Since one of our goals is
to maximise the persuasion of our system, by defining a set of user profiles it is
An Argumentation System for Assisting Users in Privacy Management 5
Fig. 1. Architecture of the argumentation system.
possible to do a generalisation of that problem depending on the user’s activity
in the social network. The information extracted to define the profiles of users is
obtained basically from three sources: the privacy configuration, the personality
analysis and the miscellaneous information.
From the privacy configuration of each user is possible to define a privacy
value that characterises him/her. Usually, a social network user defines some
privacy parameters when registering. Those parameters are the profile visibility
(e.g. public, friends, private) and the default target of his/her publications (e.g.
public, friends, group, private). With the use of the OSN this information can
be updated in order to accurately model the privacy preferences of the users.
For the personality analysis, as pointed out before, the users’ personality
can be obtained either with a survey or by analysing users activity and interac-
tions, considering the big five model [14].
Finally, when creating a profile in a social network some personal information
is added by the user (e.g. the age, the location, the likes). All this data can
also be extracted to generate the user profiles for our system. Concretely, the
miscellaneous information can be useful to determine the differences between
two different user profiles.
Apart from the users profile information, the feature extraction module also
obtains the parameters required to generate arguments in our system. Those
parameters are mainly divided into three types: the trust, the Privacy Risk
Score and the content features.
Trust is commonly understood as a way to measure the strength of a tie in
a social network. In order to formally define the trust metric we can assume a
directed graph Gt = (N , E) that represents the topology of the OSN. Let N =
{i ∈ {1, . . . , |N |}} be the set of nodes where i represents a user, and E = {(i,
j) ∈ N × N } be the set of edges representing an existing relationship from i
towards j. Therefore, the trust value ti,j indicates the strength of the tie that
links user i with user j. It is important to emphasise that, since one of the main
properties of trust is bidirectionality, the value of ti,j may differ from tj,i .
Another important parameter that the system extracts from the network
usage is the Privacy Risk Score (PRS). Proposed in [1], PRS is an alternative
way to measure the reachability of a specific user in the social network. Therefore,
the PRS provides information of the risk of sharing a determined information to
non-desired users.
6 R. Ruiz-Dolz et al.
The third parameter used in our model to build arguments is the result of
analysing the own content of the publication. The content features can come
from two main sources: the text from the publications, and the images uploaded.
We have defined the following classes of sensitive information considering the
ones proposed in [4]:
– Location: information that reveals the location of any user involved in the
interaction.
– Medical: information that reveals the medical condition of any user involved
in the interaction.
– Drug: information that reveals the use of any kind of drugs.
– Personal: information that reveals any kind of personal information. From
the sexual orientation or the job, to more identifiable information as the
credit card number, the address or the birth date of any user involved in the
interaction.
– Family/Association: information that reveals the user’s family members
or their associations.
– Political: information that reveals political orientation or any kind of ideo-
logical content.
The feature extraction module must determine whether a publication con-
tains any sensitive information or not, and classify its content in any of those
classes.
Argument generation module. The argument generation module processes
all the information gathered by the feature extraction module in order to create
abstract arguments following the guidelines of the argumentation framework.
This module generates four different types of arguments based on the type of
information extracted from the OSN:
– Privacy Arguments. This class of arguments emphasise on privacy vul-
nerabilities. The argumentation system is able to create privacy arguments
with the data obtained from the user profile modelling. The argument is
generated by computing the distance between the privacy configuration of
the user and the privacy configuration of the publication. If the distance
does not surpass a predefined threshold, the argument will have a positive
bias. On the other hand, if it surpasses the threshold the bias of the ar-
gument will be negative. Let us assume for example, a user profile with a
very restrictive privacy configuration. If that user tries to make a publica-
tion containing personal information and sharing it with all the network, the
argumentation system will create a set of arguments regarding the privacy
incoherence between the user profile and the action being done. Therefore,
the main feature to generate privacy arguments is the privacy configuration
vector of each user.
– Trust Arguments. Since the purpose of an OSN is to interact with other
users, it is a very common situation when a user involves other users with its
An Argumentation System for Assisting Users in Privacy Management 7
actions. An effective way to handle those privacy conflicts is to generate trust
arguments. An argument of trust contains all the information extracted from
the social network relative to the strength of the ties between users. These
arguments are generated from the trust computed with the information from
the feature extraction module. Concretely, if trust from users involved in
the publication towards the user making the publication is not enough (i.e.
does not surpass an established threshold), an argument of trust against
performing the action will be generated.
– Risk Arguments. When making a publication, it is impossible for the au-
thor to estimate how many users will be able to reach the information being
published. Risk arguments are generated in order to warn the user about the
risk of the publication being read by any undesired user of the network. The
main feature used to generate arguments of this type is the PRS, since the
own metric is a risk indicator. Having a high risk value will make the system
generate an argument of risk against making the publication.
– Content Arguments. Content arguments are generated from the data ob-
tained by the content features analyser. Therefore, there can be as many
content arguments as classes of content defined before. In addition, depend-
ing on the user personality and privacy configuration some types of content
arguments may be more or less persuasive. Let us suppose that there is a
user that usually shares political information on his/her posts. To warn that
user of the risks of sharing political information may have no sense. But, we
will now assume that he/she makes a post containing some sensitive med-
ical information. In this case the system will detect the risk and start an
argumentation process in order to warn the user of the risk of making the
medical information public.
The output of this module is an argumentation graph Ga = (A, R) where
arguments are the set of nodes A = { α ∈ {1, . . . , |A|}} where each α is an
argument, and edges are the relationships R = {(i, j) ∈ A × A}, where argument
i attacks argument j. Therefore, once the set of arguments is generated, the
module also creates the relationships between arguments.
In our argumentation framework, a relationship between arguments is defined
by the attack relation. To determine if there exist an attack relationship between
two different arguments, the parameter bias is used. An argument positively
biased and an argument negatively biased are both attacking each other by
definition.
Solver module. The solver module of our argumentation system performs the
task of evaluating the argumentation graph. To solve our argumentation graph,
the function τp is applied to the set of arguments generated and the profiles
involved in the argumentation process. Finally, all the scores are added and the
system checks whether the result score is positive or negative to decide the set of
acceptable arguments. If the result is positive, there are no reasons to persuade
the user on modifying his/her action. On the other hand, if the result is negative,
8 R. Ruiz-Dolz et al.
the system keeps all the negative biased arguments and sorts them by their score
in order to try to persuade the user to modify his/her action.
Dialogue module. The purpose of this module is to handle the communication
between the argumentation system and the human user. This module receives
the set of A acceptable arguments and an argumentation strategy πa . We define
an argumentation strategy as the policy (order to present the arguments) that an
argumentation agent adopts when facing an opponent (either human or agent).
The dialogue module uses each argument αn ∈ A following the strategy πa in
order to persuade the user to modify his/her action. The definition of concrete
argumentation strategies remains future work.
4 System Operation
In this section, we provide an example on how our argumentation system can be
implemented in a specific OSN to serve as an educational tool to help the users
of the network to preserve their privacy.
4.1 Pesedia: raising awareness of privacy risks in OSNs
Pesedia is an OSN for educational and research purposes that includes: (i) the
design and development of new metrics to analyze and quantify privacy risks [1];
(ii) the application of methods to change users’ behaviour regarding their privacy
concerns; (iii) the implementation of new features to improve the management
of users’ content; (iv) and the evaluation and testing of new proposals with real
users.
The underlying implementation of Pesedia uses Elgg [5], which is an open
source engine that is used to build social environments. The environment pro-
vided by this engine is similar to other social networks (e.g. Facebook). The
Pesedia architecture has two main components: the Platform Layer and the
User Layer. The Platform Layer is the core of the architecture. This layer con-
tains the Social Network Services, which provide the main functionality of the
social network, and the Storage System, which provides persistent storage of all
of the information generated in the social network. Among other modules, the
Social Network Services include the module which deploy the argumentation sys-
tem proposed in this work. The User Layer is in charge of managing information
associated to each user. This information is divided into three categories: con-
tacts (grouped or non-grouped); information (e.g., profile items, publications);
and settings, which are mainly focused on privacy settings, such as privacy poli-
cies and privacy thresholds.
4.2 Usage example
As mentioned before, a multiparty privacy conflict (MPPC) may happen when
multiple users are tagged in the same post or photo. The most common reason
An Argumentation System for Assisting Users in Privacy Management 9
for this type of conflicts are the differences between privacy policies of the users
that appear in the publication. We will now assume that a MPPC occurs in
Pesedia in order to give an example of how our argumentation system would
behave.
Thus, let us assume that user Alice wants to upload a photo to Pesedia
where also appears user Bob. Based on the features obtained by the feature ex-
traction module, the system will determine the profile information of both users
as can be seen in Table 1. In this example, user Alice main preferences regard-
ing contents are: Location(0.3), Medical(1.0), Drugs(0.8), Personal(0.5), Fam-
ily/Asociation(0.6) and Political(0.9). Meaning that when higher the value, user
Alice does less publications with that content. Values vector indicates that Alice
gives the values: privacy(0.3), popularity(1-0.3), closeness(0.5), openness(1-0.5),
flexibility(0.5), intransigence(1-0.5) and the content of the publications(0.68).
These values allow us to order the priorities of the users so that we can de-
termine the scores of the arguments for a given user. In this case, the values of
popularity, closeness and the content sensitivity are the preferred values for Alice,
so arguments regarding these issues will be more effective than others. Finally
the personality vector makes possible to model the user. In this example, Open-
ness(0.2), Conscientiousness(0.6), Extraversion(0.1), Agreableness(0.7), Neuroti-
cism(0.5); Alice can be seen as an introverted user (0.1 in extraversion) among
other parameters. This same analysis can be made equivalently with user Bob
in this example.
Alice Bob
Content preferences [0.3,1.0,0.8,0.5,0.6,0.9] [0.1,0.7,0.9,0.4,0.6,0.3]
Values [0.3,0.5,0.5,0.68] [0.9,0.7,0.7,0.56]
Personality [0.2,0.6,0.1,0.7,0.5] [0.7,0.8,0.6,0.5,0.3]
Information [21,”Aachen”] [23, ”Brussels”]
Table 1. User profile features
The module will also obtain the most important features of the content of the
post. Alice shares a photo where location and personal information is revealed,
and selects her friends as the audience to share the photo. As two users appear
in the post, the system also computes the trust between them in both directions.
In this example, the following trust values are considered: tAlice→Bob = 0.6 and
tBob→Alice = 0.3; meaning that Alice tie strength towards Bob is considerably
higher than the inverse tie. Taking into account both profiles and publication
features, the argument generation module generates the following set of argu-
ments:
10 R. Ruiz-Dolz et al.
A = {α1 (-1, T rust, LO(0.3)), α2 (+1, P rivacyA , OK(0), α3 (-1, P rivacyB ,
ER(0.4)), α4 (+1, RiskA , LO(0.23)), α5 (-1, RiskB , HI(0.78)), α6 (-1, ContentA ,
Location(0.4)),α7 (+1, ContentA , Medical(0)), α8 (+1, ContentA , Drug(0)), α9 (-
1, ContentA , Personal(0.6)), α10 (+1, ContentA , Family(0)), α11 (+1, ContentA ,
Political(0)), α12 (-1, ContentB , Location(0.4)), α13 (+1, ContentB , Medi-
cal(0)), α14 (+1, ContentB , Drug(0)), α15 (-1, ContentB , Personal(0.6)), α16 (+1,
ContentB , Family(0)), α17 (+1, ContentB , Political(0))}
The set of arguments generated contain arguments from all the four types
existing in our system. It is also possible to observe how, when more than a
user is involved in an argumentation process, some arguments are duplicated
taking into account all the different user profile features. Once all the arguments
are generated, the argument generation module also creates the argumentation
graph depicted in Figure 2. The graph generated is a bipartite graph on which
two different sets can be seen. A set with the arguments against making the
publication (top) and a set supporting the action of making the publication
(down).
Fig. 2. Argumentation graph built from the set of arguments generated.
The solver module receives the graph generated by the argument generation
module and proceeds with the argument evaluation by applying the score func-
tion to each argument and profile involved in the process. The results of the
evaluation can be observed in Table 2.
The result of aggregating all those values is -1.6 meaning that the module
will infer that the action entails some risks for both users involved in the pub-
lication. Therefore, the dialogue module will receive the set of negative biased
arguments. They will be used following a persuasion policy πAlice in order to
maximise the persuasion for that concrete user. Even though the design of the
dialogue module is still future work, we would like to illustrate how arguments
will be seen by human users in order to clearly distinguish between the shape of
arguments in the internal computation and the dialogue phase. It is important
An Argumentation System for Assisting Users in Privacy Management 11
Argument Value Argument Value Argument Value Argument Value
α1 -0.21 α6 -0.12 α11 0 α16 0
α2 0.3 α7 0 α12 -0.04 α17 0
α3 -0.36 α8 0 α13 0
α4 0.07 α9 -0.3 α14 0
α5 -0.7 α10 0 α15 -0.24
Table 2. Score obtained by each argument.
to remark that, although the system has information regarding both users Alice
and Bob, the dialogue phase will only be performed with the author of the pub-
lication, Alice in this case. Therefore, sensitive data regarding other users in the
publication will never be revealed. Let’s take into account the argument α15 , an
argument against making the publication, based on Bob’s personal content sen-
sitivity. The dialogue module will translate the argument from its computational
shape (-1, ContentB , Personal(0.6)) to a human readable text like ”It is highly
recommended not to make this publication since it may infringe Bob’s content
privacy preferences, please consider to modify the content of the publication or
to get Bob’s approval before making the publication.” where no specific sensitive
details of privacy preferences of other users are revealed.
5 Discussion
In this paper we have proposed an argumentation system to handle privacy
threats and assist users in online social networks. We have formally defined the
argumentation framework and the architecture of our system. In order to demon-
strate its operation, we have presented an usage example, where we illustrate
how the arguments are generated and how our system solves the underlying
argumentation graph.
As future work, we plan to completely integrate the proposed argumentation
system in the Pesedia network, and therefore raise awareness of the privacy
issues for the network users. We also plan to develop a chatbot to manage the
dialogue module of our system. We will implement a model based on reinforce-
ment learning able to learn argumentation strategies in order to improve the
persuasive efficiency of the system.
6 Acknowledgements
This work is partially supported by the Spanish Government project TIN2017-
89156-R, the Valencian Government project PROMETEO/2018/002, the FPI
grant BES-2015-074498 and the grant program for the recruitment of doctors
for the Spanish system of science and technology (PAID-10-14) of the Universitat
Politècnica de València.
12 R. Ruiz-Dolz et al.
References
1. Alemany, J., del Val, E., Alberola, J., Garcı́a-Fornes, A.: Estimation of privacy
risk through centrality metrics. Future Generation Computer Systems 82, 63–76
(2018)
2. Baroni, P., Rago, A., Toni, F.: How many properties do we need for gradual argu-
mentation? AAAI (2018)
3. Bench-Capon, T.: Value based argumentation frameworks. arXiv preprint
cs/0207059 (2002)
4. Caliskan Islam, A., Walsh, J., Greenstadt, R.: Privacy detective: Detecting private
information and collective privacy behavior in a large social network. In: Proceed-
ings of the 13th Workshop on Privacy in the Electronic Society. pp. 35–46. ACM
(2014)
5. Costello, C.: Elgg 1.8 social networking. Packt Publishing Ltd (2012)
6. Fogues, R.L., Murukannaiah, P.K., Such, J.M., Singh, M.P.: Sharing policies in
multiuser privacy scenarios: Incorporating context, preferences, and arguments in
decision making. ACM Transactions on Computer-Human Interaction (TOCHI)
24(1), 5 (2017)
7. Fogues, R.L., Murukanniah, P., Such, J.M., Espinosa, A., Garcia-Fornes, A., Singh,
M.: Argumentation for multi-party privacy management (2015)
8. Golbeck, J., Robles, C., Turner, K.: Predicting personality with social media. In:
CHI’11 extended abstracts on human factors in computing systems. pp. 253–262.
ACM (2011)
9. Heras, S., Atkinson, K., Botti, V., Grasso, F., Julián, V., McBurney, P.: Research
opportunities for argumentation in social networks. Artificial Intelligence Review
39(1), 39–62 (2013)
10. Heras, S., Rebollo, M., Julián, V.: A dialogue game protocol for recommendation
in social networks. In: International Workshop on Hybrid Artificial Intelligence
Systems. pp. 515–522. Springer (2008)
11. Kökciyan, N., Yaglikci, N., Yolum, P.: An argumentation approach for resolving
privacy disputes in online social networks. ACM Transactions on Internet Tech-
nology (TOIT) 17(3), 27 (2017)
12. Nute, D.: Defeasible logic. In: International Conference on Applications of Prolog.
pp. 151–169. Springer (2001)
13. Prakken, H., Sartor, G.: A dialectical model of assessing conflicting arguments in
legal reasoning. In: Logical models of legal argumentation, pp. 175–211. Springer
(1996)
14. Rothmann, S., Coetzer, E.P.: The big five personality dimensions and job perfor-
mance. SA Journal of Industrial Psychology 29(1), 68–74 (2003)
15. Schwartz, S.H.: An overview of the schwartz theory of basic values. Online readings
in Psychology and Culture 2(1), 11 (2012)
16. Subrahmanyam, K., Reich, S.M., Waechter, N., Espinoza, G.: Online and offline so-
cial networks: Use of social networking sites by emerging adults. Journal of applied
developmental psychology 29(6), 420–433 (2008)
17. Walton, D.: Argumentation theory: A very short introduction. In: Argumentation
in artificial intelligence, pp. 1–22. Springer (2009)
18. Wang, Y., Norcie, G., Komanduri, S., Acquisti, A., Leon, P.G., Cranor, L.F.: I
regretted the minute i pressed share: A qualitative study of regrets on facebook.
In: Proceedings of the seventh symposium on usable privacy and security. p. 10.
ACM (2011)