Available online at [Link].

com
Available online at [Link]
ScienceDirect
ScienceDirect
Procediaonline
Available Computer
at Science 00 (2020) 000–000
[Link]
Procedia Computer Science 00 (2020) 000–000 [Link]/locate/procedia
[Link]/locate/procedia
ScienceDirect
Procedia Computer Science 177 (2020) 178–185

The 11th International Conference on Emerging Ubiquitous Systems and Pervasive Networks
The 11th International Conference on Emerging Ubiquitous Systems and Pervasive Networks
(EUSPN 2020)
(EUSPN 2020)
November 2-5, 2020, Madeira, Portugal
November 2-5, 2020, Madeira, Portugal
Children’s Online Privacy from Parents’ Perspective
Children’s Online Privacy from Parents’ Perspective
Paweena Manotipya , Kambiz Ghazinour*
Paweena Manotipya , Kambiz Ghazinour*
Advanced Information Security and Privacy Lab, State University of New York, Canton, NY, USA
Advanced Information Security and Privacy Lab, State University of New York, Canton, NY, USA

Abstract
Abstract
A considerable number of online users are either unaware of the side-effects of sharing personal information or lack adequate
A considerable
knowledge aboutnumber of online
their rights or meansuserstoare either unaware
protecting of theThis
their privacy. side-effects of sharing
is even more personal
concerning wheninformation
we realize orthelack adequate
vulnerability
knowledge
of children’s about their rights
information or means
in social to protecting
media. their privacy.
Their information Thisposted
is often is evenonline
more byconcerning whenorwe
their parents realize thecompromising
guardians, vulnerability
of children’s
their privacy. information
In this work,inthesocial
study media. Their information
is separated is often
into two phases, posted to
a survey online by theirand
understand parents
attest or guardians,
current compromising
situation of parents’
their [Link]’s
perceptions this work,online
the study is separated
privacy, and basedintoontwo
thephases,
points alearned,
survey useto understand
a game-basedand attest
learningcurrent
in ansituation
attempt ofto parents’
educate
perceptions
parents whenofsharing
children’s
theironline privacy,
children’s and basedonline.
information on theThe
points learned,
initial usedemonstrate
findings a game-based thatlearning
althoughin parents
an attempt
and to educate
guardians
parents
might havewhen sharinglevel
a certain their of
children’s
awareness information [Link]
and knowledge The their
initialchildren’s
findings demonstrate thatsome
online privacy, although parents
of their andactions
online guardians
still
might
violatehave
their achildren’s
certain level of awareness
privacy. The results andofknowledge about their
both experiments show children’s online
that parents needprivacy, someaware
to be more of their
of online actionswhen
their actions still
violate
sharing their children’s
children’s privacy.
personal The results
information, of both
photos and experiments
videos on any show
socialthatmedia
parents need toInbeaddition,
platform. more awaremoreofneeds
their to
actions when
be done to
sharing
develop children’s personalparents
tools and educate information, photos
on methods to and videos
protect on anyonline
children’s social privacy
media platform.
as the useInofaddition, moreisneeds
social media to be done to
increasing.
develop tools and educate parents on methods to protect children’s online privacy as the use of social media is increasing.
© 2020 The Authors. Published by Elsevier B.V.
©
© 2020
2020 The
The Authors.
Authors. Published by by Elsevier B.V.
This
This is an
is an open
open accessPublished
access articleunder
article under Elsevier
theCC
the
B.V.
CCBY-NC-ND
BY-NC-NDlicense license ([Link]
([Link]
This is an open
Peer-review
Peer-review underaccess
under article under
responsibility
responsibility the
ofofthe
the CC BY-NC-ND
Conference
Conference Program
Program license ([Link]
Chairs.
Chairs.
Peer-review under responsibility of the Conference Program Chairs.
Keywords: Privacy; Children; Social Media; Parents; Survey; Game
Keywords: Privacy; Children; Social Media; Parents; Survey; Game

1. Introduction
1. Introduction
Children today have an online existence even before birth. According to a study conducted by AVG Internet
Children
Security today [1],
company havesome
an online existence
parents evenchildren’s
create their before birth. According
digital footprinttoeven
a study
at theconducted by AVG
time of their Internet
pregnancy by
Security company
sharing pictures of [1],
theirsome parentsoncreate
sonogram socialtheir children’s
media. digital with
Most mothers footprint evenunder
children at the2-year-old
time of their pregnancy
upload images by
of
sharing picturesbabies
their newborn of their sonogrammany
frequently; on social media.
of them Most
even mothersanwith
uploaded children
antenatal under
scan 2-year-old
online. upload
Besides, manyimages of
mothers
their
createnewborn
a profile babies
for theirfrequently; many of
newborn babies, themof even
some them uploaded an antenatal
give the babies scan online.
email addresses [1]. Besides, manythat
It is expected mothers
most
create a profile for their newborn babies, some of them give the babies email addresses [1]. It is expected that most
* Corresponding author. Tel.: +0-315-386-7067; E-mail address: ghazinourk@[Link]
* Corresponding author. Tel.: +0-315-386-7067; E-mail address: ghazinourk@[Link]
1877-0509 © 2020 The Authors. Published by Elsevier B.V.
1877-0509 © 2020
This is an open Thearticle
access Authors. Published
under by Elsevier B.V.
the CC BY-NC-ND license ([Link]
Peer-review
This under
is an open responsibility
access of the Conference
article under CC BY-NC-NDProgram Chairs.
license ([Link]
Peer-review under responsibility of the Conference Program Chairs.

1877-0509 © 2020 The Authors. Published by Elsevier B.V.

This is an open access article under the CC BY-NC-ND license ([Link]
Peer-review under responsibility of the Conference Program Chairs.
10.1016/[Link].2020.10.026
 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185 179
2 Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020)

children born in this digital age will have the full cycle of their digital footprint starting by the time they are 2-year-
old, continuing throughout their lifetime. It is especially important to note that the existing profile of those children
are created by parents without the children’s consent, meaning that those children do not have a chance to protect
their own privacy before the time they are mature. A study by the University of Michigan’s Health System [15]
shows that most parents share potentially embarrassing information about their children on social media. The study
also finds that about 51% of parents provide information that could lead to an identification of their children’s
locations at any given time, and about 27% of parents share potentially inappropriate photos of their children [15]. It
is crucial to recognize that some children may not desire to share such information at all when they reach adulthood,
and that children might have an interest in entering adulthood free to create their own digital footprints. Given that
such information can be potentially more harmful to children, most parents are still not aware of the consequences of
their posts, or the risks to their children’s privacy and safety.
Contributions. In this work, we first run a survey to understand parents’ perspective on their children’s online
privacy and receive 1,300 responses from parents in 51 countries. Then, based on the results of the survey, we
develop a game and use game-based learning to educate parents when sharing their children’s information online.
The rest of the paper is organized as follows: Section 2 shows the related work. Section 3 proposed our study and
its two faces. Section 4 concludes the study and discusses future research direction.

2. Related Work

There have been many attempts to propose methods to limit sharing of the Personal Identifiable Information(PII)
on social media [7, 11], visualize the privacy policies via a Privacy Policy Visualization Model (PPVM) [6] to
facilitate understanding the policies, ensuring its usability [16], or provide a platform for the negotiation between
data providers and data collectors to bridge the gap and protect their data [5]. However, in all the proposed methods,
the researchers assumed the shared information belong to the individuals who use the internet and not
parents/guardians who are in charge of their children’s personal information and this is a huge assumption to make.
The term “Sharenting” [3, 14, 15] is used to simplify the actions of parents when posting or sharing their
children’s personal information on social networks. A study by Stienburg [14] discusses an in-depth legal analysis of
the conflict between the role of parents as gatekeepers of their children’s personal information and as narrators of
their children’s personal stories. Another common case is known as “Digital Kidnapping” [14] when parents post a
few photos of their cute children, and strangers save those photos to use on their feeds as if they are their own
children. This type of digital theft can be very difficult to investigate unless someone happens to come across the
profile of that stranger. Another online stranger danger type can occur when strangers save, alter and post photos of
undressed children posted by their parents while in the bathtub, potty training, etc. [14]. The inappropriate photos of
children are not only humiliating but can also lead to the case of pedophile sites, stalkers, or even bullies. These
incidents can affect children’s mental health in the following years. Many cases exist where parents’ posts had a
negative impact on their children’s privacy and safety, such as misuse of children’s information by stranger danger,
online harassment, and cyber bullying [8].
The work in [4] discusses parents’ online behavior on posting children’s photos alone. By doing so, most parents
in the study fail to conceal related critical information such as dates of birth, full names, genders, locations, etc. In
addition, fewer parents consider the consequences of revealing their children’s faces, full names, and related
information prior to posting them. The photo sharing trend on social media is one of the largest issues, as children’s
photos have become the target of mockery and cyberbullying. Recent evidence of this can be found on Facebook
groups that collect young children’s photos and make fun of ‘ugly babies’ [15].
Social networks have both benefits and risks to children’s security and privacy, and parents need to be more
thoughtful about what information of their children they disclose online. It is true that parents who share their
children’s health condition benefit from these posts more than their children, as they get useful information, feel they
are not alone, and receive emotional support from others [13]. The downside is that some parents use their children’s
stories to express their frustrations or attempt to heal themselves. Some of them include hashtags like ‘#anxiety’ or
‘#depressed’ in their feeds to get more similar stories and expand their online community [13].
Children and adolescents are considered vulnerable groups because of their lack of maturity and life experience
[12]. For the online community, Congress created the Children’s Online Privacy Protection Act (COPPA), which
180 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185
Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020) 3

prohibits websites to collect information from people younger than 13 without parental permission. Today, most
social media providers have adopted COPPA regulations, limiting their services to users aged 13 and older.
However, many social media platforms continue to not have an age restriction, some of which are sites such as
Disney and Club Penguin which are targeted to preadolescents and younger children [12]. It is up to parents and
caregivers to evaluate and monitor social media sites for their children, as well as being active role models in using
social media services appropriately.

3. Our Study

Our proposed work is divided into two phases; the survey questions to understand and demonstrate current stages of
parents’ online behavior, and the game-based simulation to educate parents about online privacy. The survey
questions are designed to identify the level of knowledge and awareness about online privacy that parents in
different geolocations and cultures obtain, and knowledge about their children’s online behavior. Moreover, it
collected data on some similarities about parents’ online traits that might indicate parenting trends that deal with
children’s online privacy. The 24 survey questions (see [Link] for the list of questions), targeted
parents around the world with children aged 0-17, making this survey a reflection of diverse cultures and
backgrounds. Because of this, the study was anticipated to be subjective depending on the communities the
participants live in. The social media services listed on the survey are Facebook, Instagram, Snapchat, Twitter,
YouTube/YouTube Kids, Skype, Reddit, Telegram, WhatsApp, Viber, LINE, WeChat, LinkedIn, Tango, Google
products, Weibo, [Link], Pinterest, Blogger, Signal, and Game applications. These sites are chosen since they
are popular and have users around the globe. The 24 survey questions were created in the Qualtrics data collecting
platform in both multiple choice and text format. It collected data from participants by providing an anonymous link
to be posted in the Amazon Mechanical Turks (MTurks) to validate credentials and authentication of participants
and the fact that they are truly parents and guaranteed that take the survey only once. The children did not
participate in this survey. The survey received 1,300 responses from parents in 51 countries. Although there are
definitely sociological debates about cultural aspects and definition of privacy in different communities, in this
study, we do an overall presentation of our findings and plan to further investigate the sociological aspects with the
help of our expert colleagues in this field.

3.1. Phase 1: Survey Results and Findings

Based on our survey, Facebook is the number one social media service that parents around the world have accounts
for themselves and their kids. It is also a popular social media platform that most young adults use to connect with
the world, followed by YouTube/YouTube Kids, WhatsApp, Instagram, SnapChat and various gaming apps. The
top ten information that parents with children in all age groups share are 1. Gender; 2. Name/Family name; 3. Date
of Birth/age; 4. Photos/Videos of themselves; 5. E-mail address; 6. Photos/Videos of family members/pets; 7.
School name; 8. Favorite genre/sports/activities; 9. Family status; and 10. Mailing address/location/zip code.
Parents also include children’s information through their own stories in their feeds. Although the overall
responses show that parents are highly concerned about their children’s online privacy, their awareness and
knowledge on this topic is not so promising.
0-5-year-old. The parents of this age group disclose that they post and share photos, videos or information about
their child on their social media accounts, as well as tagging children’s photos and videos to family and friends. In
fact, this group of parents made up the largest group who share their children’s information on social media (73%).
Although they do not use parental control on their devices because this age group is apparently too young to browse
by themselves, many parents admit that they have no idea how to set up parental control on their devices. This issue
is also common for parents who have children in other age groups.
6-10-year-old. Most parents say they are NOT concerned about online privacy yet, although they have started to
monitor their child’s online activities or devices regularly. According to the responses from parents in this age
group, some children start to register and request to be members of game applications and social media services like
YouTube/YouTube Kids, WhatsApp, Signal, and [Link], and have their own social media accounts. Therefore,
parents have started to monitor their children’s online activities regularly and have set up parental control on the
devices because they find it is too dangerous to leave their children using social media services without supervision.
 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185 181
4 Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020)

In contrast, parents rarely discuss the dangers of being online with children at this age group, simply because they
assume that children are too young to understand it. In addition, parents do not know if their child’s information
appears elsewhere without their consent. They also reveal that they must seek information from their child’s
network to find out if the photos, videos or information were misused, as their children do not share this information
directly with them. In this age group, parents begin to receive peer pressure to open social media accounts for their
children, sharing and tagging photos and videos on their accounts with family members and friends.
Although most parents have experience using social media platforms, their knowledge on social media
functionalities is surprisingly inadequate, especially on the options to ‘opt-out ‘and ‘remove’ personal information.
However, parents gain more experience and become more aware of how to remove personal information or how to
set parental control on devices when their kids reach the ages in which they can browse the internet by themselves.
11-13-year-old. Parents show the most concern about online privacy (60%) compared to other age groups. Both
parents and children have their own personal devices (mostly tablets and mobile phones) making it very challenging
to monitor activities frequently. Also, parents of kids in this age group continue to have peer pressure on using
social media services to share their children’s stories. On the positive side, parents set up parental control on their
children’s devices, and monitor and talk to their children regularly about the dangers of social media. Most parents
with children in this age group care about their children’s rights and privacy and believe that it is the right thing to
set parental control for particular websites and devices so that they are not exposed to inappropriate content.
However, many parents in this group do not know how to do set parental control and admit that they trust their
children because their children know more about the internet than they do!
14-17-year old. These trends continue to occur when adolescents reach age 14-17. Unfortunately, it shows that
parents do not have the knowledge when they enter parenthood and have not been willing to obtain this knowledge
as their children grow. Another aspect of children’s accessibility to online content shows the potential risk of
children’s online privacy violation. Children in this age group start to show demands on using certain social media
services and online applications, as the majority of parents report that their children have access to their online
account and/or credit card (e.g. iTunes, PayPal) for purchasing games and music. Parents with children in this age
group show concern about their children’s privacy; however, they do not interfere as much with their children’s
online activities as those in younger age groups. As children reach young adulthood, parents do not view their social
media activities regularly anymore and the children no longer need parents to open social media accounts for them.
The survey shows although parents do not monitor their children’s every activity online, they still kept an eye on
some activities such as what their children post, as well as liked and tagged photos and videos of themselves. As
some parents become friends in their children’s networks, they automatically monitor some online activities which
they think are enough and appropriate for keeping privacy between parents and young adult children. In addition,
parents with children in this age group shared children’s personal information less than those in an earlier age group.

Issues on awareness and knowledge about online privacy. From the parents’ perspective, online privacy in social
media services is very subjective and confusing, especially between data owner and the data collector [2]. Although
most parents state that they care about online privacy (when asked at the very beginning of the survey), they also say
they do not think social media services request too much information from their users. Moreover, when asked at the
very end of the survey ‘Do they think/believe there is such a thing as online privacy?’, the majority of parents admit
that no such online privacy exists, or they do not know if it has ever existed. This perception can be interpreted that
the parents today may have a lack of understanding about the definition of online privacy, or that they do not have
enough knowledge to clarify what kind of information is potentially harmful to their own privacy as well as that of
their children. Most parents (73%) of all age groups reveal that they post/share photos, videos or information about
their child online, and that almost 90% of those parents share them on their own social media accounts. Many
parents also consider the popularity of social media sites over ratings/reviews when signing up for the services. In
addition, they do not consider privacy on dissemination of user information prior to signing up for the service at all,
as fewer parents care about data dissemination policy and read the social network’s privacy policies.
It is worth considering that many parents seem to be very positive about online experiences because the majority
of parents (62%) never receive any complaint about related online activities from their children, and never receive
notifications or reports that their children’s information, photos, or videos has been used elsewhere without their
consent. However, out of those who indicate that they received complaints, most of these came from an
overabundance of advertisements, spam and fraud. Interestingly, among those complaints, children also experience
cyber-bullying more and more as they grow up with the children between 14 and 17 have the highest percentage.
 Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020) 5
182 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185

Sensitive Data Disclosure. Based on the survey participants’ feedback, we see the information they shared by
parents (and their children) over the internet contains sensitive PII, such as names and family names, dates of birth,
genders and e-mail addresses. It is more likely that users who share this information with a high visibility level can
be identified easily and face security and privacy issues. The survey results reveal the fact that some of the
information come from the posters’ first legal documents such as school ID, driving license, etc. is very alarming.
While parents share their own stories online as well as those of their children, most parents state their concerns
over items over which they could not control and shared by their children, themselves. When asked ‘what changes
would you suggest to social media services regarding security and privacy policies?’, the majority of parents at all
age groups (57%) prefer social media service providers to allow them to have access to their children’s posts and
chat history. 55% of the parents also share concerns about the purpose of the data used in the online world, as they
prefer social media sites to have a clear statement on how their information or their children’s information is going
to be used and would like the privacy policy to be written in a clear and concise manner so that they can easily read
and understand it before collecting data. Possibly this is one of the main reasons they do not read the privacy
policies in the first place. Interestingly only 29% of parents share concerns on data retention and the fact that they do
not know how long their data or their children’s data would remain in the social network application’s database.

3.2. Phase 2: Game-Based Learning Approach

The findings from the first phase of the survey demonstrated the need for parents to be more aware of their online
behaviors and the lack of knowledge on protecting their own privacy and their children’s privacy. The second phase
is focused on advocating parents using a game simulation to mimic their online environment, showing the
consequences of their behavior using a score-based model. The goal of this simulation is to educate parents through
direct experience when they choose to share sensitive data and photos that have potential risk of privacy violation,
including reasons why those choices could lead to negative outcomes. Social Sim Parents game is designed to
examine the parents’ awareness and knowledge about information shared on their social networks. The user
interface of the game is like a social network web interface. However, features are designed to facilitate users with
simple tasks and a few descriptions about each proposed scenario. This online web-based game has 6 parts:
a) Welcome page, prompts participant on the information about the purpose of this research, credentials, and how to
start the game; b)Profile page, contains query boxes that the participants can choose to fill in; c)Scenarios, contains
5 steps/scenarios (see Table 1), for participants to choose how they post contents, along with the photo in each one;
d)Favorite page, where the participants select subjects that they and their children like (sports, shopping, vacation,
etc.) e)Score card, provides feedback for participants’ performance on each step with the number of total likes, total
scores, reasons, and privacy risk level; f)Post-survey, records feedbacks from the participants after playing.
The game is written in JavaScript language and is available at [Link]. All photos used in this game
come from the internet (mainly from Google and Bing images). As shown in Figure 1, there are two parts of data
collection: 1) The data from the game such as the identifier data in the profile page, posts, reasons for selecting the
photos, and scores; 2) The post-survey to see the participants’ level of awareness and knowledge. The database to
store data from the game is created using Google Spreadsheets and the post-survey is via Qualtrics platform.

Fig 1. The architecture of Social Sim Parents game

 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185 183
6 Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020)

This web-based, mock-up social network game is designed to be a simple game with simple rules. Each
participant is given the initial score of 100 points which will be reduced throughout the game depending on the
amount and sensitivity of the information shared. The game has a total of 52 query fields, including 16 query boxes
on the sign-up page, 20 opt-in query boxes on the favorite page, and 16 clicks and posts on the scenarios/steps 1-5.
The game starts by prompting the player the Welcome page that instructs the participants to sign up prior to
playing the game. Once the player clicks Sign-up, it prompts participants to the Profile page where they need to
enter information in order to move on to the next step. The query boxes on this page are optional, with the
exceptions of the e-mail address and password that need to be filled in to sign-in and access the game and are
marked with a red star on the corner of the boxes. The query boxes/fields with the sensitive identifier data (e.g., date
of birth and address) and non-sensitive ones are not distinguished to see how much personal data participants agree
to share and how many boxes they leave blank. The more they share, the more negative scores they get (-1 point for
any box/field they fill). To protect the privacy of participants we do not store the actual answers and cannot verify if
they enter real or fake data. For simplicity we assume they enter true data. If the participant decides to put
information only in the mandatory fields (with red stars on the top right corner of the query box), their points remain
unchanged. If the password contains less than eight characters, it is considered not secure and points are deducted.
The game starts and the participants are given 5 steps that get more severe as they progress, see Table 1, and
asked to pick one of the photos suggested by the game and post a content with it. They are also asked to explain why
they picked the photo and should set a privacy setting for the post that makes it visible to the public, the social
network, or only themselves.

Table 1. The five scenarios given to participants to post and select a photo with them
STEPS SCENARIOS

STEP- You just bought a new house and would like to tell your networks about your good news. Your kids also just start
1 their first day at school or day-care.

STEP- Your child reaches the age that needs to be toilet trained. You also want to teach your child other disciplines since
2 the child has started to disobey your house rules and disrespects others in public space.

STEP- Your kid receives some achievements (getting driving license/ accepted to a nice school) and you want to share the
3 news with others. You are very happy and want to reward your kid and family by taking them on a nice vacation)

STEP- You learn that your child starts to complain about going to school. Their academic performance also drops
4 dramatically. You suspect that your child may be bullied (or cyberbullied) or has learning difficulties, but you are
not so certain about the causes.

STEP- You find out that your child is diagnosed with Learning Disabilities and ADHD. You must bring your child to
5 receive therapy once a week.

After the 5 steps are over, the participants are taken to the Favorite page, where they pick their favorite subjects
(e.g. shopping, movies, sports, etc.) for themselves and their children. Choosing the subjects in this step is
completely optional and we would like to see how much information the participants are willing to disclose. For
every subject selected they receive a – 1 point.
After this step, the participants are taken to the Score card page where the see their scores and feedback on why
they got the score. For example, if they posted the picture of a kid holding his driver’s license to share the news that
their kid got his license, they get a negative score and the game explains that the kid’s face and their legal document
can be seen on the page that is visible to the public.
As the last step of the game, the participants are taken to the post-survey page to give feedback about the game.
Scoring. The scores are measured based on the granularity level [2] of the data revealed in the posts, the privacy
setting of the post which shows the visibility level of the post, and the severity of the steps of the game (1-5). Step 5
potentially reveals more sensitive information compared to Step 1. The number of Likes generated for a post is
based on both visibility and granularity level. The Like score is used as a positive reinforcement for the participants
to share data with high visibility level (public and friends) to find more friends or followers, along with the
184 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185
Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020) 7

likelihood of having photos chosen with positive feedback. In contrast, players will receive a Like score of -1 if they
set the visibility level of their posts or photos only to be seen by themselves.
In order to have a proper score for the content of the post, text-based analysis is used. As this research intends to
evaluate the appropriateness of contents that parents choose to describe each scenario that relates to their children, it
is essential to have a system to detect positive and negative language used in the posts. The Sentiment Analysis
dictionary [10] is the tool used in this experiment to identify positive and negative English words using Natural
Language Processing to show scores in each step. Most posts are short sentences with clear single words and fewer
misclassifications of sentences or words. This is important, as some English words can have multiple meanings in
different contexts. For example, the word ‘unpredictable’ is positive in the context of the thriller genre, but negative
in other contexts. Using the Sentiment Analysis dictionary is among the best practices although it may not be as
accurate as desired. For simplicity, we give +1 point for every positive word (e.g. excellent, great, happy) and -1
point for every negative word in the post (e.g., sad, disappointing, lazy).

3.3. Game Results and Key Findings.

From the profile page, 60% of participants provided a complete Date of Birth which is considered a high-risk
behavior for identifying an individual. However, some of them chose to provide only partial data such as only the
month and year, or date and year; this reduced the risk level. In addition, 22% of them chose not to provide their
date of birth at all. Another key identifiable data that negatively affected scores was the address; unfortunately, 64%
of the participants agreed to provide their complete address with their zip code in the query boxes.
In addition, among the 16 query boxes on the sign-up page, 80% of the participants in this study provided input in
every field, even though they were optional. On the privacy setting page, 44% of parents in this study restricted their
visibility level to only me, while 40% set their profile as public, and the rest chose to friends.
Most of participants in this game (61%) were highly concerned over their privacy and their children’s privacy
(low risk), while only 8% fell into the high-risk category, and 31% were in the medium risk range. While scoring
was tied with many attributes, the number of Likes was tied solely with the level of visibility (public, friends, only
me). 44% of the participants set their visibility level on both profile and posts as only me, making the number of
Likes lower than expected. This can be explained as an awareness of the game’s purpose: to examine their
knowledge about online privacy; therefore, participants were more cautious and thus restricted their privacy settings
to be less visible and accessible compared to their real-life behaviour.

3.4. Post-Survey Results

46% of the participants in this experiment had children aged 0-5, followed by those with children aged 6-10 (34%),
11-13, and 14-17, respectively. 65% of the participants in the game declared that their knowledge about online
privacy was ‘very good’ prior to playing the game. A small percentage of parents in this experiment claimed their
knowledge was either ‘excellent’ or ‘poor’ (16%). However, the number of parents who claimed their knowledge on
this topic dramatically increased to ‘excellent’ after they played the game rise from 16% to 27%. No one claimed
that their knowledge was ‘poor’ after finishing the simulation.
When asked about their level of awareness about online privacy reflected in their score cards, many parents stated
that they were aware of those potential risks ‘regularly’. Most of these parents paid very close attention to the
‘Personal Information Disclosure’ issue while they were performing their tasks, rather than worrying about the
number of ‘Likes’ received. However, 18% of the participants disclosed that scoring affected their photo choices
and motivation to finish the task. Other than the topic on personal information disclosure and scoring, some parents
shared concerns over their post contents and photo choices. However, when addressing the overall satisfaction from
the learning perspective, most parents said that they were ‘satisfied’ (43%), and ‘very satisfied’ (22%). Very few
participants needed more instructions and mentioned they were ‘confused’ or ‘could do better’. Also, participants
indicated that they learned it is possible to opt-out information from social media networks after playing the game.
From the two experiments, there are some important findings that need to be addressed in this study when
addressing the domain of social media users’ behavior and their thinking processes. Firstly, most participants
successfully raised their levels of awareness about online privacy after participating in the social media simulation
and most of them were considering changing their online behavior when they post information on social media.
Secondly, sometimes people did not post what was relevant to the photo. This is because some participants
 Paweena Manotipya et al. / Procedia Computer Science 177 (2020) 178–185 185
8 Kambiz Ghazinour et al./ International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2020)

provided reasons when they picked a photo for a post, they chose pictures based on what fits with their expectation.
For example, a participant mentioned ‘This one is positive and gives good feelings to my friends’, ‘It's good
photography and powerful’, or ‘ To let friends & family know that my child starts new school so they can wish her
well.’ In addition, some of them even chose random photos that reminded them of their past experiences rather than
a photo that is related to the post (or the given scenarios). Moreover, people can post things with hidden messages or
implied meanings such as metaphors, poems, sarcasm, or idioms that need to be interpreted. Therefore, posting
contents can be very subjective or seemingly irrelevant.
Thirdly, from the learning outcome perspective, parents in this study did not demonstrate significant change of
their knowledge about oversharing personal information after playing the game. From the sign-up page to the end,
participants still had learning curves on how to remove and opt-out existing information that they agreed to share,
even when provided with a reminder statement. This can be due to their unfamiliarity with the user interface,
learners’ limitations, or unclear instruction. Overall, most participants provided positive feedback for this study and
we hope it encourages researchers to perform further research on the topic of children’s online privacy from parents’
perspective.

4. Conclusion and Future Work

The Social Sim Parents game is in its early stages. The design of user interface, game flow, instruction, variety of
photo choices, and aesthetic aspects can be modified and improved to better stimulate the social media environment
and encourage the participants to be more engaged in the tasks and feel entertained at the same time.
There are some risk factors, or indicators such as stolen identity, fraud, or life safety that can be referred to as
high-risk; however, there are many types of photos and contents in grey areas that could be categorized as medium
risk. Such a categorization method can be challenging for accurate measurement. The Sentiment Analysis dictionary
also has limitations on accuracy when used in various contexts, as the automated tool still in the process of
development, the Sentiment Analysis database will hopefully continue to be improved upon with time.

References
[1] AVG Technology. 2010. Digital Birth: Welcome to the Online World. In Business Wire. DOI:
[Link]/news/home/20101006006722/en/Digital-Birth-Online-World
[2] Ken Barker, Mina Askari, Mishtu Banerjee, Kambiz Ghazinour, Brenan Mackas, Maryam Majedi, Sampson Pun, and Adepele Williams.
2009. A data privacy taxonomy. In British National Conference on Databases. Springer, Berlin, Heidelberg.
[3] Lisa Belkin. 2013. Humiliating Children in Public: A New Parenting Trend? In HUFFINGTON POST.
[Link]
[4] Janice Gatti. 2009. A thin line: 2009 AP-TVT digital abuse study. In MTV. DOI:
[Link]/MTVAP_Digital_Abuse_Study_Executive_Summary. pdf.
[5] Kambiz Ghazinour, Ken Barker. A privacy preserving model bridging data provider and collector preferences. In the Proceedings of the Joint
EDBT/ICDT 2013 Workshops. March 2013. Genoa, Italy. pp. 174-178.
[6] Kambiz Ghazinour, Maryam Majedi, Ken Barker. A model for privacy policy visualization. In the proceedings of the 33rd Annual IEEE
International Computer Software and Applications Conference (COMPSAC). July 2009. Seattle, USA. pp. 335-340.
[7] Kambiz Ghazinour, Stan Matwin, Marina Sokolova. Monitoring and recommending privacy settings in social networks. Proceedings of the
Joint EDBT/ICDT 2013 Workshops, March 2013. Genoa, Italy. pp. 164-168.
[8] Tara Haelle. 2016. Do Parents Invade Children’s Privacy When They Post Photos Online? In NPR. DOI: [Link] npr. org/sections/
health-shots/2016/10/28/499595298/do-parents-invade-childrens-privacy-when-they-post-photos-online
[9] Lauren S Jordan. 2016. Writing for the Mighty, for My Son, and with My Son. Washington Post. DOI :
[Link]
[10] Bing Liu. 2012. Sentiment analysis and opinion mining. In Synthesis lectures on human language technologies. 5.1, 1-167.
[11] Maryam Majedi, Kambiz Ghazinour, Amir H. Chinaei, Ken Barker. SQL privacy model for social networks. In the Proceedings of the 2009
IEEE International Conference on Advances in Social Network Analysis and Mining (ASONAM), 20-22 July, Greece. 2 pages.
[12] Gwenn S O'Keeffe, and Kathleen Clarke-Pearson. 2011. The impact of social media on children, adolescents, and families. In
Pediatrics .127.4 : 800-804.
[13] Amy Quintal. 2016. When You Need to Help Yourself Before You Help Your Child. In THE MIGHTY. DOI :
[Link] child/
[14] Stacey. B Steinberg. 2017. Sharenting: Children's Privacy in the Age of Social Media. In Emory Law Journal, 66(4), 839-884.
[15] University of Michigan Health System. 2015. Sharenting' trends: Do parents share too much about their kids on social media? In
ScienceDaily. DOI: [Link]/releases/2015/03/[Link]
[16] Tahani Albalawi and Kambiz Ghazinour, “A usability study on the privacy policy visualization model,” in IEEE 14th Intl Conf on
Dependable, Autonomic and Secure Computing, Auckland, New Zealand, August 8-12, 2016, pp. 578–585.