This is a version of a publication
in
Please cite the publication as follows:
DOI:
Copyright of the original publication:
This is a parallel published version of an original publication.
This version can differ from the original published article.
published by
‘Unmochon’: A Tool to Combat Online Sexual Harassment over
Facebook Messenger
Sultana Sharifa, Deb Mitrasree, Bhattacharjee Ananya, Hasan Shaid, Alam S.M.
Raihanul, Chakraborty Trishna, Roy Prianka, Ahmed Samira Fairuz, Moitra
Aparna, Amin M Ashraful, Islam A.K.M. Najmul, Ahmed Syed Ishtiaque
Sultana Sharifa, Deb Mitrasree, Bhattacharjee Ananya, Hasan Shaid, Alam S.M. Raihanul,
Chakraborty Trishna, Roy Prianka, Ahmed Samira Fairuz, Moitra Aparna, Amin M Ashraful, Islam
A.K.M. Najmul, Ahmed Syed Ishtiaque. (2021). ‘Unmochon’: A Tool to Combat Online Sexual
Harassment over Facebook Messenger. CHI '21: Proceedings of the 2021 CHI Conference on
Human Factors in Computing Systems. May 2021. Article No.: 707.pp. 1–18.
DOI: /10.1145/3411764.3445154
Author's accepted manuscript (AAM)
Association for Computing Machinery
CHI '21: Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems
/10.1145/3411764.3445154
© 2021 ACM, Inc.
‘Unmochon’: A Tool to Combat Online Sexual Harassment over
Facebook Messenger
Women in the global south often seek justice of online harassment through unavailing the harassers and the screenshots of
their sent harassment texts and visual contents on social media. However, women often experience further backlash with
the argument that these screenshots may have been manipulated. We studied these issues through a survey with 91 female
Facebook users and interviews with 43 females and other stakeholders from Bangladesh and designed ‘Unmochon’- a tool that
helps women collect such harassment texts and visual contents from Facebook Messenger, share with their intended group of
users and prove their authenticity. Our user-study with 48 participants revealed various challenging aspects of seeking justice
on social media using such technologies and the assumptions they are built on. Based these findings, we further discuss how
technologies should be designed to address women’s harassment on social media in such a complex social setting.
CCS Concepts: •Computer systems organization→ Embedded systems; Redundancy; Robotics; •Networks→Network
reliability; • Online Gender Harassment, Social Media, Facebook, Messenger, Solidarity, Feminism, Bangladesh;
Additional Key Words and Phrases: Online Gender Harassment, Social Media, Facebook, Messenger, Solidarity, Feminism,
Bangladesh, Unmochon, User-evaluation
1 INTRODUCTION
Women’s harassment is one of the age-old problems in the history of humankind. Many international reports
explain how prevalent this problem is from the enlightened western world to low-resourced global south. For
example, the United Nation Women reported that Up to 70 percent of women across the world face physical
and/or sexual violence in their lifetime [7] and National Intimate Partner and Sexual Violence Survey (NIPSV)
reported in 2018 that about 1 out of every 5 women (21.3%) in the United States experienced at least one attempted
or completed rape [82]. The picture in the global south is also equally disturbing, if not worse, as 66% of women
reported experiencing sexual harassment between two and five times in Delhi, India [7].
Today, with the widespread advancement of information technology, this problem has taken over the cyber-
spaces. Thus, online gender harassment is a widely discussed agenda across the world. Prior studies have showed
that female social accounts experience a significantly larger number of harassment incidents than that of male
accounts [9, 21, 29, 62] and thus have concluded that women are more vulnerable to online gender harassment.
The situation is more severe in patriarchal societies like Bangladesh as reports show that more than 70% percent
of the complaints filed at the government’s Information and Communication Technology Division’s Cyber Help
Desk in 2017 was about women’s online harassment [50]. Such harassment incidents often stress out the victims
and their families and lead to further effects like exclusion, humiliation, and public resentment, especially when
they live in conservative and patriarchal societies [55]. However, harassers often remain beyond the scope of
law-enforcement as Mishuk Chakma, Additional Deputy Commissioner of the Cyber Security and Crime Division
of Dhaka Metropolitan Police (DMP), commented, “However, most (female) victims of online harassment do not go
to the police station" [26].
Among the very tiny portion of the female victims of online gender harassment who break the silence and seek
out for help, often prefers exposing the harassers on social media with screenshots of their chats as supporting
evidence and seek justice [47, 52]. But, in most cases, the reports and the press releases by concerned authorities
show that the victims’ attempt to seek help and justice go in vain with the argument that the advancement of
technology now allows manipulation of screenshots. Thus, offenders often go away receiving the benefit of
doubts [47, 52]. However, any formal research has not reported or clarified yet what level of the extremeness of
the harassment pushes them to take such steps and what still blocks them from seeking out help.
1
We address this gap in the literature through a two-phase span study. In the first phase, we investigated the
above-mentioned queries through a survey with 91 female Facebook users and interviews with 43 Facebook
users. Over the course of this phase, we sought answers to the following research questions:
RQ1: What kind of gender harassment do Bangladeshi women often face online, and which kind of
challenges often block them from seeking help?
RQ2: When and why do Bangladeshi women prefer exposing their online harassers and what factors
are significant while choosing possible online spaces for such actions?
RQ3: How do the existing tools, platform policies, and online environment often fail to assist them in
this regard?
Our findings from this first phase show that often their complaints about harassment and supporting evidence
are challenged with authenticity. Building on these, the second phase of the study aimed at designing a tool to
support the women who seek help and justice to their online harassment. We build on the shame-based model
of justice in this adversarial design [12, 28]. In this regard, we prototyped an application for women to help
them seek justice for their online harassment. This paper introduces ‘Unmochon’, an application that helps with
authentication issues associated with the screenshots of the chat history with the harasser that women often
share while seeking help and justice. We evaluated this application with 48 participants using interviews and
focus group discussions. We found that women often prefer hiding their identity as a victim, fearing that this
might bring them further victim-blaming and slut-shaming and may even engender real-life security threats.
During the user-evaluation of the tool, some participants also mentioned that any misuse of such technologies
could lead to counter-harassment for the accused party and urged to include such concerns in the set of agendas
while designing feminist technologies.
Our work makes a four-fold contribution to feminist-human computer interaction (feminist-HCI), social
media studies, and information and communication technologies and development (ICTD). First, our findings
detail Bangladeshi female internet users’ online gender harassment and what kind of measures do they take
while addressing those. Second, drawing on our findings, we build a set of design assumptions practical and
appropriate for assisting victims to seek justice. Here, we bring the model of shame-based justice to design [12].
Prior works suggest that societies may have lower crimes, if shame about the crime is communicated effectively
[16]. Therefore, shame about harassing women must be communicated effectively for seeking online gender
Justice and reducing women harassment. In this regard, we prototype and evaluate Unmochon and report the
perception of victims and other stakeholders of such tools that support online gender justice. Finally, we report
findings from our discussions with the participants and other stakeholders that unfold different complicated
issues entangled with cyber-harassment and social justice and suggest possible design directions for safer online
experience for women in the global south.
2 RELATED WORK
In building our research question and making arguments based on our findings, we draw on the literature of three
broader themes: technologies supporting women’s safety, online movements for gender justice, and literature of
justice and gender harassment in Bangladesh. In this section, we briefly discuss some of the existing works built
on these themes and situate our research question in this regard.
2.1 Gender Harassment, Cyber Spaces, and Related Technologies
Online harassment is a widely discussed agenda in HCI, social media research, and gender studies. Online
harassment not only stresses the users out but also may cause economic damage to the platform when users
choose to leave it or limit their activity [33, 59, 90]. Even though most of the online platforms and social media
offer the users with help to handle such cases, often users find those processes slow, inconsistent, and dismissive
2
Unmochon
[40, 91, 92]. Researchers working in the intersection of social media, web security, and HCI design have been
engaged in developingmachine learningmodels to detect harassment for years [12, 20, 36, 57, 70, 97]. Some of these
models have also integrated human intervention in cases for further effectiveness of the models [12, 20, 36, 70, 97].
However, these models often fail to address the needs of the users due to a higher odds of deception and
misclassification, and the presence of biases in the process of their training mechanisms [11, 46]. Along with
all of the above-mentioned concerns and limitations, the existing state-of-the-art tools and mechanisms for
fighting online harassment often fail to address culturally contextual gendered needs in such cases of harassment,
including users’ vulnerability due to their gender identity and a balance between their needs and expectation and
their power and agency on social media [12, 33, 80, 84, 90].
In recent years, a group of researchers has showed that users’ perception of online safety is subjective and
depends on a lot of factors including their digital privacy and security and the values upheld by the community
[73]. Some researchers have also engaged in understanding forms of harassment, its gendered angles, sealioning,
and other relevant abuse on social media and call for designing more options in the platforms than blocking
harassers and deactivating own IDs in this regard [12, 13, 76, 90]. Jhaver et al. studied Twitter users’ blocklists to
interrogate online harassment and reported that users may not feel entirely protected by blocking the harassers
and often blocked users feel they are unfairly treated [51]. However, the design, features, and affordances of social
media, including Facebook, vary across regions (see Facebook lite and messenger lite for the global south [78])
and thus users’ social media behavior are also diverse and culture-specific [61, 98]. Thus, the aforementioned
rich body of works needs to expand by investigating how the patterns of harassment vary for users of different
populations and how the existing supporting tools and techniques are challenged by those.
Another body of literature has progressed over time in explaining how internet connectivity, messenger
services, multimedia calls, and social media have been useful for women addressing and fighting their harassment
and abuse. For example, Satchell and Foth found that women on physical spaces and on-street pretended to
talk or text on their phones so that the prospective harassers nearby would think that the women are active on
social network right at that moment and they have the option to seek help right away if needed [77]. To boost
their courage against prospective street-harassers, some women also reported preferring to remain connected
to their loved ones over the time of their travel and inform their loved ones over a phone call upon arriving at
their destination [14]. Furthermore, there are many designs and interventions to address harassment on streets
or other public places in western countries. They include ComfortZones [14], CityWatch [53], Hollaback! [27],
CampusWatch [71] among others.
However, addressing gender harassment using technology is more challenging in low-resourced countries in
the global south because many of the users live in strictly patriarchal societies, and raising voice against men is
often challenging and usually discouraged through victim-blaming and slut-shaming [3, 56, 86, 88]. Still, a thread
of such design has evolved over time to address harassment and crime in the physical space through design. For
example, Protibadi, ProtibadiNext, Safestreet, and HearMe in Bangladesh [2–5], Harassmap in Egypt [99], Safe
Mathare in Kenya [39], CrimeID in Indonesia [96], and Safetipin, Samrisa, and Panic-button in India [56, 83, 89]
were designed to address street harassment, seek help from passers-by, friends, or police, share stories online,
and/or map harassment incidents on an interactive digital map. However, gender harassment on social media,
messenger services, and cyberspaces in the context of the global south has remained largely an understudied
topic in HCI, feminist-design, social media research, and related domains. One notable exception is Nova et al.’s
work on understanding the pattern of sexual harassment on social media in Bangladesh [68], where they have
revealed how people close to women harass them anonymously using the social media. However, their work
does not involve any preventive design interventions to stop such harassment. This calls for growing a deeper
understanding of the types and ways of harassment that Bangladeshi women often face online, and
which kind of challenges often block them from seeking help (RQ1) so that we can design a safer online
space for them.
3
2.2 Online Gender Movements and Protests
The history of online movements and protests is fairly new but rich. Often such movements and protests come
with keywords that have specific meaning followed by hashtags and some activities. Covering all of them is
beyond the scope of the paper. Here, we will briefly describe some of the major and relevant feminist online
movements and protests.
Many of the early online feminist movements were launched basing on ideological conflicts. For example, Pink
Hijab Day (PHD) in 2004 was related to breast cancer awareness and aimed at countering the stereotypes of
Muslim women [75]. The movement started in the US and gradually spread through social media calling for
solidarity. However, this movement did not receive much organizational support outside of the US, and many
Arab women denied their solidarity because of local constraints. Another significant Muslim women solidarity
movement also took place in the last decade in Germany when a gender activist organization, namely FEMEN,
aimed at gender-freedom and called for several on-street activities that did not align with core sentiment of many
of the Muslim women [93]. Those groups of Muslim women, self-defines as ‘Muslima Pride’, protested on street
and spread the movement through Facebook, Twitter, and Instagram. using hashtag #MuslimahPride.
Black feminist agendas hold a significant part of online movements when it comes to protesting against
injustice. Most of those lies in the intersection of racism and gendered crimes. For example, #SayHerName started
in 2015 to tell the story of violence against black women while circumventing traditional media barriers [95]. It
started with the idea of making the sexual assaults and deaths of black transgender women noticeable which the
mainstream media kept ignoring, bringing out the stories of police brutality and violence against black women,
and gaining justice for them. Similarly, #JadaPose, #StandWithJada, and #JusticeForJada on twitter and other
social media in 2014 stood for soliciting justice for Jada, a sixteen-year-old rape-victim black teen [94]. Another
significant movement, #JusticeforLiz, shook the internet and social media in 2013 and sought justice for Kenyan
teen rape victim pseudonymously tagged as ‘Liz’ [45].
Arguably the biggest online movement broke the internet in October 2017, in which women across the world
used the hashtag #MeToo (or some variants of it) and publicly shared their untold stories of being sexually
harassed [81]. Founded in 2006 by Tarana Burke, a black female activist from New York [17, 18, 81], these events
followed Hollywood actress Alyssa Milano’s call on Twitter: “If all the women who have been sexually harassed
or assaulted wrote ‘Me too’ as a status, we might give people a sense of the magnitude of the problem" on October
15, 2017 [38]. This call for sharing harassment experiences with #MeToo hashtag followed Milano’s own story
of being sexually abused [63] and millions of women joined this movement. Some literature sees #MeToo as a
descendent of prior #MyHarveyWeinstein, #YouOkSis, and #SurvivorPrivilege movements [42, 66] and argue
that it only gained prominence when several Hollywood female celebrities came out with their stories of sexual
harassment [79].
Within a few days, #MeToo reached the Indian subcontinent, as many women flooded Twitter and other social
media platforms with stories of their sexual harassment [43, 58, 66]. However, this wave also faced a similar
backlash that victims here felt discomfort in disclosing stories of harassment if their harassers were someone in
the upper level of the power structure. For example, Sarkar, a Dalit Queer, cohered with #MeToo by preparing
a list of harassers in Indian Academia [35]. In this regard, Sarkar shared that the victims didn’t provide any
context while reporting and didn’t share any details explaining the incidents. The official investigation found a
few academicians guilty with no news of ongoing or planned inquiries for other names in the list [24, 79]. A
retired female showbiz celebrity shook the media in late September 2018 with her posts about the experience of
harassment by a colleague [25, 60]. Many other female and queer colleagues joined her quickly breaking their
silence [79]. While this movement successfully triggered the academia, industry, and showbiz, the harassment of
a large group of women from other professions, class, and social status remained beyond its scope. Also, reports
4
Unmochon
show that #MeToo in Bangladesh was less successful than expected because of cultural differences, lack of hope,
and a lacking reliance on alternatives [43, 66].
The limited success of #MeToo in the Indian subcontinent aligns with the previous pattern of online movement
of #SlutWalk back in 2012, which also flopped in this region [65]. Following the western wave of fighting for
women’s right to choose their attire and not to be slut-shammed [64], there was a call for #SlutWalk in Bhopal,
India. This was officially declared as a failure due to insufficient in-person participation, while thousands of
people showed their interest on Facebook and other social media platforms and promised to join [65]. However,
not all the online movements stay unsuccessful in the Indian subcontinent. For example, protest against increased
tax on female hygiene products [#LahukaLagaan, see [30]] is known to be one of the biggest women-led Twitter
campaigns to demand policy change in India. This provokes us to investigate how social media and other
technologies together leverage gender justice in the Indian subcontinent and we ask what kind of roles do
different tools and technologies play or underplay in this regard (RQ3).
2.3 Model of Justice and Gender Harassment in Bangladesh
Finally, we turn to the literature of justice relevant to our context and research question. Prior works on gender
justice showed that many female victims of street harassment are reluctant to report to police and law enforcement
authorities since those may not have been much unhelpful previously and the investigations may lead to the
next level of inconvenience and costing another harassment to themselves [48]. Also, reports show that women
often feel that their reports to law-enforcement authorities may also bring them shame and blame for supposedly
provocative behavior [48]. Another study revealed that such incidents often leave great stress on women, resulting
in making them feel helpless, and a further limit on their mobility [3]. Also, they are generally not comfortable
engaging in discussions regarding such harassment incidents because there is some social backlash of bringing
attention to street harassment, and disclosure of them being harassed could further itself be played as a source of
shame and embarrassment [3] for the female victims. It has been reported that women often vent their frustration
regarding such incidents in social media and disclose their experience to friends, peers, or followers [6, 27]
and there are tools and interventions that support such agenda of gender justice sought online by disclosing
identifiers, an image of the face of the harasser, for example [27].
The reports on seeking justice online in the context of Bangladeshi gender harassment also cohere with this
pattern, as we see from the reports [26, 50]. In most of the cases, women are fed up with the response from
their situated infrastructure of legal and social support and thus find alternatives by exposing and shaming the
harassers in front of internet-citizens, often termed as ‘Netizens’. This model of seeking justice aligns with the
idea of ‘Mob Justice’ .Mob Justice is defined as, “Ruling by a group of mass" [41, 54]. Here the Netizens constitute
the mob who decide the conviction. Although part of such a justice model seems to be unpredictable public
opinion, mob justice has often been differentiated as a model of justice when proper democratic participation is
missing in a justice system and people feel in-secured and lose their faith in legal systems [69]. This practice
of Bangladeshi women seeking mob justice excites one of our research questions. In this study, we investigate
when and why do they prefer exposing their online harassers and what factors are significant while
choosing possible online spaces for such actions (RQ2).
3 UNDERSTANDING ONLINE GENDER HARASSMENT IN BANGLADESH
We conducted an anonymous survey and investigated the nature of women’s experiences of online gender
harassment, their existing strategies for seeking help and justice for it, and how their networks often react to
that. Then, through interviews with 43 participants, we grew a deeper understanding of various angles of online
gender harassment to find possible design avenue. This phase of work was conducted between June to October
2019. This study was approved by the ethics review committee of the authors’ institutions.
5
Details of Pre-Design Survey and Interview
Survey
Total Number of Survey Responses: 91
Age Range (in Years) Occupation
18-25: 52 (57%) Student: 53 (58%)
26-35: 36 (40%) Public/private job-holder: 27 (29%)
36-45: 2 (2%) Teacher: 5 (6%)
46+: 1 (1%) Others: 6 (7%)
Interview
Total Number of Interview Participants: 43 (Female:37, Male:6)
Age Range (in Years) Occupation
18-25: 18 (42%) Public/private job-holder: 27 (63%)
26-35: 23 (53%) Student: 7 (16%)
36-45: 2 (5%) Housewife: 3 (7%)
46+: 0 (0%) Others: 6 (14%)
Table 1. Demographic Characteristics of the Survey and Interview Participants
3.1 Methods
To get a broader understanding of Bangladeshi women’s harassment experiences online and to develop ideas
of their resistance and associated challenges, we conducted both an online survey and an interview study. The
anonymous online survey allowed many women to talk about this tabooed issue more comfortably. On the
other hand, the interviews allowed us to go deeper into some of the stories coming from some women who felt
comfortable to talk about this issue with us. We describe both of our methods below with further details.
3.1.1 Online Survey. We hosted our anonymous online survey on Google Survey and circulated it online,
including emails of the authors’ friends and Facebook groups which the authors’ frequently used. Participants
took part in the online survey by clicking on a link attached to the emails or the Facebook post. We requested a
response from Bangladeshi female users only. The request for participants’ consent came along with the post
and the email. In the call for a response to the survey, we explained the purpose of the survey and ensured the
anonymity of the respondents. The participants were allowed to proceed upon confirming at the beginning that
this was their first time participating in the survey.
The language of the survey was Bengali. It asked the participants about the type of harassment they faced
on Facebook and other social media platforms, their relationship with the harassers, if they sought help from
friends and law-enforcement authorities and other relevant information. Although most of the questions were
structured as check-boxes or multiple-choice questions, the survey also included several optional open-ended
text-boxes where participants could freely express their opinions and concerns around their harassment on
Facebook messenger and other social platforms. No personally identifiable information (e.g., Name, Location.
etc.) was collected from the survey participants in this study. We did not offer any compensation for participation.
It generally took the participants around 10-15 minutes to complete the survey. The participants were allowed
to leave the survey at any time without any further consequences just by closing the survey window in their
computer browser. A total of 91 participants completed the survey. See table-1 for the age range and occupation
of the participants.
3.1.2 Interviews. To grow a deeper understanding of different angles of online gender harassment, we conducted
semi-structured interviews with 37 female and 6 male Facebook users. Among them, there were 3 male police
6
Unmochon
officers who have been working for several years on gender harassment on cyber-spaces and one professional
female lawyer who has been working on gender issues for the past three years. We recruited the participants
from our professional networks on Facebook through convenient sampling [49]. We explained the purpose of the
work to them while scheduling the one-on-one interviews. Since gender harassment is a stigmatized topic in
Bengali culture, we let them choose between online and in-person meetings.
We conducted the interviews in Bengali. We asked them about their and their friends’ experiences of online
harassment, especially in comments and chats on social media, what kind of actions they took, what kind of
help they sought, and other relevant information. We led to the deeper discussions only when they confirmed
us of their comfort in talking about those incidents. The Police officers and the lawyer helped us with different
socio-cultural, ethical, and legal angles of online gender harassment and existing support systems in the country
for the victims. All the interviews started with oral consent from the participants and were audio-recorded in
researchers’ mobile phones with the permission of the participants. We also took notes during the interviews.
Generally, the interview sessions lasted around 30-40mins.
3.1.3 Data Collection and Analysis. All the survey responses were initially recorded in the storage of the Google
survey. The data came anonymized and we retrieved them in comma-separated values (CSV) format. Later we
stored it in our secured storage space. First, we translated the CSV from Bengali to English and cleaned the
data. We used the open source statistical tools on python for further analysis. We also evaluated the responses
qualitatively to understand the nuance of the online gender harassment faced by Bangladeshi female social media
users.
The interviews generated a total of 11 hrs of audio recordings. We transferred them to a secured computer
owned by the researchers. Then we translated and transcribed them. The transcriptions and the interview
notes generated 80 pages of documented data. We removed the identifiers before conducting open coding and
thematic analysis on them [15, 85]. Four of the authors independently read through the transcripts carefully
and allowed codes to develop. Later they shared their codes with each other. A Total of Thirty-five codes were
spontaneously developed initially, for example, stigma, victim-blaming, skepticism and mistrust, denial, slut-
shaming, betrayal, blackmail, threat, etc. After a few iterations, we clustered related codes into themes and drew
our design assumptions on them.
3.2 Understanding Online Gender Harassment in Bangladesh
The findings from the survey and the interviews inform us about the harassers and the type of harassment often
the victims face, effects of harassment on them, existing social and legal supports and how those are helpful to
them, and how they often resist. This subsection highlights some of the key findings.
3.2.1 The Harassers. We found that more than 70 of our survey respondents (58%) were harassed by unknown
people online [see fig-1(left)]. 11 of the interview participants mentioned that they tried visiting the Facebook
profiles of their unfamiliar harassers but their information was hidden or their Facebook profile used "locked"
feature. However, unknown online IDs may not be always strangers as 5 of the interview participants mentioned
that they had clues and evidence that their unknown harassers were someone they knew. We quote one of them,
“The unknown ID in the Facebook other-box shamed me for my bigger breasts and my skin being dark.
Later we found that a group of students from our class together ran that account. Rather than apologizing
for what they did, they said, "Relax, chill, and take it easy. It was not to insult you, we were just kidding."
But that was no fun for me!", (P15)
45 of the survey respondents (36%) said that their harassers were their classmates, friends, and relatives who did
not even bother hiding and harassed them upfront [see fig-1(left)]. A survey participant expressed her frustration
and intimidation as her harasser was a law student at her university and hoped that this survey might help her
7
Fig. 1. (Left) Relationship between the harassers and female victims in context of harassment over Facebook Messenger,
(right) if women consider seeking legal support and why, if not
find ways. Another interview participant mentioned that her husband lived overseas and her sister-in-law and
sister-in-law’s husband are the temporary guardians of her toddler and her. She explained how her sister-in-law’s
husband tried to take advantage of her and used Facebook messenger for that,
“The man keeps texting me on Facebook asking to see him outside alone, sends me vulgar images, and
deletes those from his end in a few minutes. I can not even block him because then what will I tell my
sister-in-law? They live next door, I often need to leave my baby to them if I go outside. I can not offend
her, right?", (P28)
Another 13 of the interview participants shared similar stories of their online harassment incidents and
explained how those were challenging since those came from family members or close relatives. Five of them
even refused to discuss the details of the incidents as they tagged those as “private problem within the family" and
thought discussing those would be inappropriate even for research purposes.
3.2.2 Types of Harassment. Our survey reports that online harassers often send women vulgar text messages,
and inappropriate multimedia contents, request nude photos of them, threaten them of defaming using their
photo-shopped pornographic images, and even sometimes threaten them to rape and murder [see fig-1(right) for
the breakdown]. We divide the patterns of harassment in major five categories: forcing a romantic relationship,
seeking revenge, spreading hatred and threatening, and venting random frustration.
Forcing for a Romantic Relationship. 39 of the survey respondents reported that they were forced to go in
a romantic relationship, date them, or marry them [see fig-1(right)]. Sometimes such proposals come with even
worse kinds of threat, as 11 of the survey participants mentioned in their response. We quote one of the survey
participants’ explanations form her response in the survey,
“One of my friends was threatened (on Facebook) by a senior student (who wanted to date her) at (name
of the university where she studies), he said, “I will kidnap you from the university campus if you do not
meet me in person". She was extremely frightened.", (S80)
15 of our interview participants informed us that they were repetitively asked them out and marry them
through texts and calls, and even in comment sections of their and their friends’ posts by the harassers.
8
Unmochon
Seeking Revenge. 13 of our survey respondents mentioned that their harassers were their ex-boyfriends or
someone whom they have refused to date or marry before. Two of the respondent mentioned that their harasser
continued even though the respondents are married for several years. Often these harassers tried to defame them
in public, as we quote one of them,
“(After I refused to date my classmate)...he told people the truth by mixing lies in my name and tried to
gain sympathy. He wanted to portray that, it’s me who played dirty with him, and threatened me that
he will reveal all the things I share with him when we used to be friends and will prove to everyone that I
am bad and so on...", (S34)
Thus, harassers of this category are generally known to the victims and often they have something to defame
the victims, including romantic chat history, previous intimate photos, and other private multimedia contents.
Imposing Conservative and Religious Sentiments. 19 of the female interview participants mentioned that
their harassers, both online and in real life, used conservative and religious sentiments to harass them. In many
cases, often the female victims are cursed for numerous reasons including their working or studying outside
and their attire as a justification for their harassment. Our participants opined that online spaces are easier for
imposing these. Even our survey received such a response that name-called one of our male researchers and
expressed objection against his wife’s attire, as we copy that response,
“Girls need to cover themselves properly by doing “Parda". That will help to lessen harassment. People
along with (One of the male co-author’s name) and others, please keep your wives covered following the
rules of “Parda".", (S27)
Mockery Using Anti-Religious Sentiments. 4 of our interview participants were Hindu females, and all of
them mentioned that their online harassers frequently used mockery and disrespectful words towards Hindu
religion. They explained that being women of the minor religious group makes them easier targets for such
harassment, as we quote one of them,
“Being a Hindu is already challenging here, being a Hindu female is even worse... One Facebook account
sent me some awful vulgar texts about Draupadi and her five husbands (characters from Hindu religious
myths). It said that they were also five people running that account and kept asking me out with
inappropriate offers... Using my religion to harass me is twice as hurtful.", (P15)
Venting Out Random Frustration. Our survey informs that 17 participants’ harassers called them over the
phone or messenger calls services and initiated threatening and inappropriate conversations. 21 participants
received threatening and erotic texts, or vulgar messages, inappropriate audio clips, and multimedia contents
from unknown accounts. Harassment for no reason may also come from known people, as one of the participants
gave an example of her friend’s experience,
“... Another friend of mine was consoling a guy she knew after his breakup, and he started saying things
like ‘I named my dick the hulk’ and then sent her pics (images). It was so uncalled for, considering we
knew him as an elite person.", (S80)
11 of our female interview participants also mentioned that they thought most of the harassers they experienced
were possibly frustrated and harassing random women online was probably a reflection of that. They also opined
that harassers probably think women’s inboxes are the safest places for venting out their random frustrations.
3.2.3 Effects of Harassment on Victims. In many cases, women’s harassment is considered as normal phenomena
by their friends and family and other possible support. Sometimes their complaints and frustrations are even
discarded as a topic of discussion as if pretty women are supposed to be harassed, as one of the survey participants
explained,
9
Fig. 2. (a) Break down of the forms of harassment that female Facebook users face over the messenger inbox
“...(T)he comment was like, “As a beautiful girl, it’s pretty common that guys will fall for you, so it’s
very normal. When pretty girls are given freedom, the girl and her family members/husband need to
accept this kind of situations and understand that these kinds of incidents can happen with her.", (S60)
Furthermore, when a female participant discovered that she never experienced any gender harassment, not
that she was aware of and she expressed her surprise in the survey response. In the survey and during the
interview, many participants mentioned that taking those was never easy for them and these incidents made
some emotional, social, and educational damage, as we discuss them below.
Restricting Social Life.Often womenwho face harassment are ashamed in society, as many of our participants
informed us. 14 of our participants mentioned that in many cases their friendly behavior was misread as sexual
signaling and they were harassed. As one of the participants explained by giving an example of her experience,
“I was involved in a lot of organizational activities. But then some students started saying that I am
doing those for attention. One day a text from a co-organizer arrived on my phone asking me how many
times he needs to use his (body-part) so that I am convinced to be his partner in some event. When I saw
him on campus, I asked him and he said he did not do it. But I know either he or his friends have done
that. I have stopped being social since then, I do events occasionally, but I have a cold face now.", (P19)
Our interview participants also mentioned that this fact equally applies to domestic environments as well. 14
of the interview participants reported that at some point in their lives, they were harassed by their family or
next-door neighbors. One of the participants shared her experience of being harassed by a cousin and how their
families went into a bitter relationship,
“My cousin who is 8 years younger than me used to text me frequently and I replied nicely. Then he
started sending me inappropriate media contents. I was shocked and asked him not to send those to me.
Rather than stopping, he called me old-fashioned. I blocked him because he won’t listen to me. Later my
aunt, my mom, and I had an unpleasant conversation and my aunt cursed me. Now my mother blames
me for speaking up. But I am not going to be social with that family anymore.", (P32)
Thus, often women stop being social in their real-life to avoid harassment. This sometimes helps and sometimes
fails, as several participants mention. Furthermore, sometimes they also need to limit their online social life
because of their harassment, as 9 of the women mentioned and here we quote one of them.
“I was first harassed online on Mig chat. I was only 17 then and it was only a few months that I started
using it. An ID wanted me to act like I am doing physical intercourse with him, I did not even know that
10
Unmochon
called sexting. I was scared and stopped using it. But that was a way for me to stay connected to a lot of
friends, so I lost those connections.", (P19)
Limiting Scope of Education and Career. Our participants mentioned that things are worse when their
harassers share the same academic community. 13 of the interview participants mentioned that they have
experienced harassment from their classmates and some of them had to compromise their academic scope and
opportunities because of this issue, as we quote one of them,
“I used to ask a lot of questions in the Facebook group of the class. That person put likes on all my
comments. Then he texted me asking to see him separately and help him study. He would never stop
nagging so I blocked him. Then he confronted me in the class one day. I was scared. I went to the
register-office immediately, changed my class shift to morning ones. Also, I deactivated my Facebook
so that there is no way he can reach out to me, not even using another Facebook ID. But this is not a
sustainable solution. It is very easy to find me on campus and harass me again.", (P23)
However, things are more complicated when the harasser is somewhere in an upper level of the power structure.
In such situation, the victims are thought to be making things up to climb the ladder and further become victim
of slut-shaming. 15 of the interview participants mentioned either them or their friends have become the victim
of such slut-shaming, as we quote one of them,
“He is a junior teacher in my department at the university. He seemed very friendly with the boys in my
class but he texted me strange things using Facebook message. I was astonished. I did not know whom to
talk to, so when I saw my advisor next time, I informed him and asked him what to do. My advisor said
that I should stop flirting with him as he would be beyond my league. But I did not flirt with him! And
beyond my league!? What league? He was the awful one!", (P27)
9 of our participants also mentioned that at some point of their lives, their parents thought of stopping them
from going to school, tutoring center, or for education in general because they were victims of harassers from
their schools or tutoring centers and avoiding those places was the best solution their parents thought of. 5 of
the participants also mentioned that their mothers were child-brides as their grandparents married their mothers
off to avoid the trouble of eve-teasing on their way to school.
Furthermore, 13 of our participants reported that their or their friends’ professional career was impacted
because of online harassment at some point. One of them explained,
“My friend used to work at a USA based foreign company and her boss was located in the USA. So, often
she had to work in the night shift and keep talking to her boss over messenger. He used to send her adult
contents jokingly and later explained that such contents would keep her awake at night. And he used to
send those in the messenger’s secret conversation mode, so those would disappear in a few seconds. My
friend told him that she is uncomfortable. But nothing worked. My friend just stopped going to the office,
stopped receiving their calls, and did not even go there to retrieve her experience letter.", (P9)
Another 3 female participants mentioned that they also changed their job because of their harassment that
took place within their office network and internal platform and they found no other way than quitting.
Emotional Damage. 21 of our interview participants termed online harassment as ‘stressful’. They mentioned
how such incidents stopped them from concentrating on their work momentarily and for a longer time. Even if
they tried working around it, things did not go back to normal as before, as one on the participants explained
this way,
“It is not easy to stay calm once someone you do not know threats you that they will enter in the night
through your window and rape you. How do I even know that I do not know them? What if they are
someone who has access to my house? When I first received a threat like this in my other box, I could not
sleep a whole night.", (P22)
11
However, 7 women mentioned in their interviews that the harasser being a known person was difficult for
them to handle and that is how it stressed them out. We quote one of them,
“One of my male friends in the class sent me awful jokes in the middle of one night. I thought he must
have been sleepy while sending those. I ignored it and pretended that I did not see since talking about
those would initiate further awkwardness. But things got worse when a male friend of him cracked a
joke of me being his sexting partner. That was untrue and humiliating in front of a bunch of classmates.
No matter how much I explained, I received that sarcastic gaze from them that I am one slutty bitch. I
have stopped making friends. If friends are supposed to not support you rather make fun of you then I do
not need one.", (P31)
Sometimes, even if the victims are classmates, their lack of control over the environment gives the harassers
some extra benefit, and this further damages the victims emotionally. One of the participants explained how it
was difficult for them, as we quote her below,
“I was one of the friendliest persons, but not anymore. They misinterpreted me intentionally and further
posted my caring comments and questions as I am trying to play slut with them. Seeing a strange vulgar
framing of my questions and comments as Facebook posts was traumatic for me. I begged them to stop,
but they did not listen and said, “it is just a joke". I was so broken that I had to take sleeping pills for the
next few nights. I had to take sessions (at the psychiatrist’s).", (P17)
When we asked her if she took any step against it, she mentioned that she made a complaint at the school but
her school refused to look into that case and told her that cyber-space is beyond the scope of the school campus.
3.2.4 Existing Social and Legal Support. In the survey, we asked if they ever considered seeking help from
police and law-enforcement authorities. 19 of them responded that they did not know the appropriate process.
Another 10 mentioned in their survey responses that their parents are reluctant to get into legal issues. Some of
the respondents mentioned their mistrust in local law and law enforcement agencies, as we quote one of them,
“After I reported the incident to one of the faculties, he told me to contact the police authority. But I
didn’t dare to speak about this at home because my family is a little backdated, so they will blame me
instead. And I have enough doubts about whether the police will take it seriously if I tell them about this
incident.", (S34)
One of the survey participants also mentioned in her response that involving the law-enforcement agencies
might even bring her further trouble in the day-to-day life security, as we quote her,
“I do not know if it can be brought under the law. And if there is a law which is not enforced, I am not
sure whether it will be a threat to my security or not.", (S2)
When we discussed this with three of our professionals, they explained to us that many times it is difficult to
find the harassers and bring them under the law because they either use a fake Facebook ID, or use cloned ID.
Also, they mentioned the challenge with digital evidence,
“Often we receive complaints about online gender harassment and they mention of comments and private
chats. They also send us a copy of those in images, but those images may not work as solid evidence. See,
today, there are many software that would manipulate the images of chats and comment on Facebook.
One way to solve this challenge is that they would come by our office, log in to their Facebook from our
computer, and show us the evidence from there. Or, they can log in using a desktop and the browser to
enter Facebook and video record the chat history along with the address bar on the top so that we can see
who exactly was the person they chat with. This might minimize the scope of manipulation.", (P41)
12
Unmochon
However, in many cases these are troubles for many of the victims because of their limited resources. Also,
filing police cases takes a long time to solve. Thus, many of our participants mentioned that even if they know
the process, they would just pass complaining.
3.2.5 Resistance. During the sessions, we investigated how do the participants address their online harassment
and documented their response. We cluster the ways in three themes: ignoring and blocking, reporting to the
authority and the platform, and expose and shame. We briefly discuss them below:
Ignore and Block. 27 of our interview participants informed us that they think it is better to ignore the harasser
silently. 75 of our survey respondents and 23 of the interview participants mentioned that they have blocked
their harassers on the social media they used. But blocking does not end the story in many cases, especially when
the harassers are desperate and fearless. As one of the interview participants explained,
“I had a friend who was harassed by a man from her village. He used to text her asking out and my
friends never agreed. He sent awful things and he will do those to her once he can force marry her. No
matter how many times she blocked his phone number on her phone, and his Facebook IDs, it does not
help. He would change the phone number and get a new Facebook ID and restart.", (P9)
Another participant gave an example of her friend who was harassed by her school teacher in class and also
over Imo a few years back. The participant explained how the blocking and ignoring was not feasible for her, as
we quote her,
“He used to touch from the side while pretending to check her papers. In the night he would explain in
Imo that how much fun it was to touch her breast that way. My friend could not tell her parents because
he was the teacher, could not even block him because her parents are connected to him using Imo.", (P7)
Thus, when women are harassed by known people and have connections in real life, they try blocking and
ignoring but those do not solve their problems due to their social relationship.
Report to Platforms and Rightful Authorities. 23 of our interview respondents mentioned that they have
reported their unknown harassers on the social media platform they used. But the process of reporting to social
media platform often also suggests them to block that person, which is not a practical solution in many cases as we
explained earlier. Furthermore, online harassment in professional groups is more challenging for women to handle
this way. Over the course of the interview, we collected 13 cases where women’s online harassment happened in
their workspace, and reporting to rightful authority was difficult since those harassers were either their seniors
at work or their favored ones, as 11 of the stories suggest. In the rest of 2 cases, when the victims complained to
their higher officials, those cases were resolved with minimum punishment, as one of the respondents shared her
friend’s experience,
“Her teammate wanted to date her and she was not interested. He started cursing her in texts and even in
the office-emails since she ignored him. When she reported to her team-lead with copies of those emails,
he ‘helped’ resolving the case and said, ‘This is a minor problem and this may happen when we work
together. Let it go.’ Their office asked her to delete the copies of the emails from her end since they wanted
to keep the environment toxicity free and keep the reputation of the team neat and clean.", (P12)
Thus, this participant explained her friend’s frustration as she could not block or ignore him, neither her
reporting to the rightful authority worked, and she was requested to destroy the evidence.
3.2.6 Motivation to Expose and Shame. Our participants mentioned several reasons that influence online
gender harassment victims to expose their harassers online. These reasons include: i) many of the victims find
justice in shaming the harassers in public, ii) women want to warn others to be aware of the harassers, iii)
exposing the harassers online help them prove the case of harassment, and iv) in many cases, exposing helps to
push the formal reporting to the rightful authorities. We discuss these reasons below.
13
Justice in Shaming. Sometimes women facing harassment feel that exposing their harassers should be a
possible response while seeking gender justice. 21 of the interview participants mentioned that they think shaming
is the way to prevail when it is challenging to bring the harassers under justice through social, professional, and
legal systems. One of the participants who lived in a small town shared a story of a school teacher who harassed
her and many other girls over messengers.
“He picked the under-performing girls, sent them porn images, and asked them to explain the details. If
someone refused, then he would complain to her parents that she is inconsistent with studies. One of the
girls finally exposed his stupid activities. I am glad that what I could not do during my days, she did it.
Now he is known as a predator in the town, he has lost most of his private tuition. Girls like me were
scared and suffered alone, he should suffer in public.", (P37)
Thus, 15 of interview participants mentioned that they have considered exposing or have exposed their harasser
by publishing the harassment stories and a copy of their conversations, if available, in public. The victims believed
that this would impact on harassers’ social and professional life and that is the punishment they deserve.
Blowing the Whistle to Warn Others. 9 of the interview participants who went to the same university
mentioned a similar story of a harasser at the university who harassed many female students in their Facebook
inbox and later got exposed to a Facebook group related to that university. One of the participants explained the
story to us,
“My roommate was a victim of his consistent dating proposal. And a girl living 2 rooms after ours also
kept receiving the same proposals. My friend was afraid that the guy might get vindictive. Meanwhile,
his harassment stories started coming out in (name of the group). My friend and I convinced a male
friend to share screenshots from our side to show that there are more victims of this harasser. And we
wanted to tell the other female students that they should be aware of this person.", (P29)
Thus, these female participants mentioned that they did not see prevailing justice only through punishment or
shaming, but through finding a way that no person becomes a victim of the same thing again. They also explained
that since society is patriarchal and men are considered superior here, it is an injustice to a girl not to inform her
about the predators around them.
Burden of Proof. 13 of our participants opined that when a female victim raised their voice against their
repetitive harassers, often they are asked to submit proofs. In many cases, their cases are dismissed because they
do not save those or post those. One of the participants explained this by sharing her experience with us. She
explained that she was involved in organizing an event with a bunch of senior alumni at her school and one of
them harassed her online. He kept asking her to see him alone and spend some ‘quality time’ as he termed that
way. When she raised her voice against it, she was shut because of a lack of proof, as she explained,
“He used to sendme strange texts using secret messagemode (on FacebookMessenger). Once the celebration
was over, I raised my voice and posted the story with a copy of the most recent text from him on the
school alumni group. But his friends asked me where is the proof that he ridiculed me consistently and
why am I saying it now... I wish I saved the previous texts as well.", (P23)
The participant further explained that how the friends of that harasser started dogpiling. Another 5 participants
also shared a similar pattern of such experiences when people asked for proofs even for repetitive harassment
incidents.
Exposing is Reporting. As we showed that sometimes formally reporting to the concerned authorities failed
and dismissed, organizing mini-movements on social media against the harassers might help establish gender
justice. For example, here is a story that one of the interview participants shared. A renowned professor in
their department was accused of online gender harassment. Initially, some of his students and colleagues denied
accepting that because of his skyrocketing reputation. We quote one of the interview participants,
14
Unmochon
“When the first girl raised her voice on Facebook, they challenged her by asking, “Did he really harass you
in the chat?". Then she gave screenshots and they said she made it up. Then gradually more women came
out, some from other universities and even school kids who got connected to him through a competition.
Some girls claimed that they formally reported at the university but their cases were dismissed as he is a
star professor. Finally, when he lost his reputation, the university stepped in and fired him.", (P20)
Our participants mentioned that sometimes it is difficult even to start a conversation on harassment due to
fear of further victim-blaming and slut-shamming. Thus, 17 women mentioned that while reporting these to
anyone outside their friends, women should either use a fake ID, possibly as male one, to hide behind it and avoid
victim-blaming and slut-shaming, or request a male friend to speak for them for more credibility.
3.3 Summary and What is Missing
In summary, we found that women are often victim-blamed and slut-shamed for their harassment. Therefore,
they generally prefer to keep harassment incidents a secret. They would share within their friends and solicit
possible solutions. Sometimes, they also try to locally solve it while hiding their identity. The existing process of
seeking legal help is often complicated, lengthy, and challenging for them due to a number of reasons. So they
respond to their harassment by exposing the harasser with evidence of the history of conversation and showing
solidarity to other victims if harassed by the same harasser. Still, they often face backlash from the people they
seek support from as these groups of people often deny to believe the incidents of harassment. In such cases,
even if the victims present the screenshots, often people are skeptical if that evidence is authentic.
The professionals working for police informed us that confirming the authenticity of digital evidence of
harassment is challenging and the forensic procedure is also complicated. Still, they try to work it out by either
in-person evidence collection, or videos of chat where they actually get to see the harassers’ ID. However, this
complicated process further discourages the victims to seek legal help. Today, this part of the design of social
media is poor. Widely used social media platforms in Bangladesh, including Facebook, are missing a feature
which would help online gender harassment victim prove the authenticity of their harassment evidence and help
them engage in conversation with the law enforcement organizations to establish online gender justice with
minimum hassle.
4 DESIGNING ‘UNMOCHON’: A TOOL SUPPORTING GENDER JUSTICE
We draw on the findings from the survey responses and the interviews and translated those into the design.
We named our tool Unmochon. It is a Bengali word which means ‘disclosing’ or ‘unavailing’ something. One of
our pre-design interview participants discussed an idea of a hypothetical anti-harassment tool and named that
‘Unmochon’. Later, we borrowed it from them. This section describes our design goals and components of the
application followed by the details of implementation.
4.1 Design Goals
Our findings from the pre-design survey and interviews inform us of the denial that the victim receives while
seeking online gender justice. Therefore, we set our objective to help women collect evidence of their harassment
in a way that could be accepted as authentic by the concerned ones. We borrow the idea of shaming from shame-
based design for gender justice [12]. We select Facebook as the platform for the deployment since Facebook is
one of the most widely used social media platforms in Bangladesh. We draw on our findings to determine four
design goals necessary for a successful online anti-gender harassment tool:
G1: Capture Screenshot with Authenticity. The participants described scenarios where they saw
general people on Facebook being suspicious of the screenshots being true. In some cases, the
15
Fig. 3. Work-flow diagram of Unmochon application with three major components: plugin at the user-end, the server with
end-to-end encryption, and the customized Facebook group.
screenshots were manipulated and some words were replaced. This kind of scenario urge us to set
the goal to design a tool that captures the screenshot with minimum or no option of moderation.
G2: Confirm that the Facebook Account in the Chat and Facebook Account Being Accused is
the Same. In some cases, someone opened a cloned Facebook ID of a particular person, pretended
to be him, and harassed women over the chat to defame him. This scenario urges to set the goal to
design a tool that confirms that Facebook account in the chat and Facebook account being accused
is the same. This will also make sure that no innocent individual gets reported and shamed in this
process.
G3: Hide the Identity of the Victims. Some of the participants reported that they would prefer
opening another Facebook account or requesting friends to make the case on behalf of them so that
they would stay behind the curtain and avoid victim-blaming and slut-shaming. Thus, our third goal
of the design is to create such an affordance and lessen women’s dependency on others.
G4: Share Screenshot in Public. In most cases, victims share the evidence of harassment in public
to shame the harasser, find more victims of the same harasser and share moral support, and seek mob
justice. Our final goal is built on this existing practice and we aim to create a convenient platform for
them to share their evidence and proof of its authenticity with minimum hassle.
16
Unmochon
4.2 Components and Workflow of Unmochon
From the user needs and design goals arising from the interviews, we prototyped Unmochon. It is a windows-based
application. It has three main components: the browser plugin at the user-end, a server for storage, and a dedicated
Facebook group run by human admins.
4.2.1 The Browser Plugin. The application comes with an unmochon.exe plugin. The purpose of the plugin is to
take the screenshots from the chat window of the victim and prepare it for sharing. The initial version of the
plugin that we used for the study only works on the Chrome browser. It is developed on Java Platform.
Taking Screenshot. Once the plugin is installed, the welcome window pops up with a button on it saying
‘Take Screenshot’. After pressing the button, the application will allow the users with 10 seconds to set up their
Facebook messenger chat thread on the Chrome browser of which they want to capture the screenshot. Once the
10 seconds is over, it will automatically capture a screenshot and copy the URL from the address bar that shows
the unique Facebook ID number of the account with which the user had a chat and took a screenshot of.
Hiding Identifiers. Upon capturing the screenshot, the application opens the editing window that allows the
users to use the drawing pen only so that they can hide their own name, display picture in the thumb and any
related sensitive information, which they prefer not to the public (see the components and the flow in fig-3 (left)).
Once they are done hiding such identifiers, the application allows them to save the image of the chat and set it
ready for sending it to the server for further process. However, the step of hiding identifiers is optional and users
can skip it by pressing ‘cancel’ button and proceed to the next step of reporting by sending it and the harasser’s
Facebook account number to the server.
4.2.2 Storage in Server. Once the screenshots reach the server, it stores them there. Then it puts a stamp on
the image that says ‘Verified by Unmochon’ which means the screenshot of harassment chat history was taken
and reported using the application and thus it is possibly not fabricated. The mechanism of putting the stamp is
preset and functions only when the Facebook account number reported by the user and the one retrieved by
the plugin while taking the screenshot is the same (see the components and the flow in fig-3(top-right)). The
mechanism also detects the mode of the reports by analyzing if the user has kept their identifiers (mode-1) or
hidden those (mode-2) in the sent image, before sending it to the Facebook group. Thus a package of four items
is sent to the Facebook group, the image, the victim’s requested mode and Facebook account number, accused
harasser’s Facebook account number, and how many times the same Facebook account number has previously
been reported as a harasser. We have used google cloud platform for the server. All the communication between
the server and the Facebook users are end-to-end encrypted.
4.2.3 Facebook Group. This is a Facebook group with all the posts preset to the privacy setting of ‘public’. Only
the application developed by us can post here upon being approved by human admins (see the components and
the flow in fig-3 (bottom-right)). This admin panel consists of responsible individuals of the society including
members of law enforcement, psychologists, and gender activists. Once the package from the server arrives at
the group and requests posting, the admins check if the report is a spam. There could be two types of mode
of reporting. If it is ‘mode-1’, then the post comes with the victim’s name and Facebook account number, the
screenshot of the harassment chat-history, the accused harasser’s Facebook account number, and how many
times this person has been reported previously. For a ‘mode-2’ reporting, the post discards the victim’s name
and Facebook account. We have used Facebook Graph API for the communication between the server and the
Facebook group.
5 USER-EVALUATION OF UNMOCHON
Our findings from the survey and pre-design interviews suggest that online harassment of women is a sensitive
issue and often victims are in a vulnerable position due to lack of socio-cultural support. So, we were wary of
17
Fig. 4. Unmochon application takes a screenshot of Chrome tab that is currently open. It also copies the harasser’s unique
Facebook id from the address-bar. Before sending it to the database and posting to the dedicated Facebook page for Unmochon,
the application window only allows the users to mark the photo with red ink. (All the purple geometric shapes in the figure
were post-edited for anonymization.)
conducting an intervention with the participants, rather we decided to share our plugin with the users with
no active contact with the server and the Facebook group, explain the whole idea, and seek their feedback. We
first prepare a user-study package, which included a prototype version of the plugin and the user-guide to the
application. To avoid unintended posting during the user-study, we shared the version of the prototype with no
connection to the server and the Facebook group (see fig-4, attempt to posting on Facebook failed). For their
better understanding, we also created a demo video and added that to the package that we shared with the
participants. This section details the methods used and the data collection and analysis process.
5.1 Methods Used to Collect User-Feedback
5.1.1 Focus Group Discussion (FGD). We conducted a total of 6 FGD sessions with 19 female (4 groups) and
10 male (2 groups) participants. 13 of the female participants were from the set of participants of pre-design
interviews. We recruited the rest of the participants from our professional networks on social media, from
Facebook groups and friends. Upon reaching out to them, we explained the purpose of the work. We grouped
relatively known participants of the same gender together and set up FGDs with them. We also sent them
the user-study package a few hours before the interview so that they experience those and get some time to
think about those. In the FGDs, we explained the purpose of the application and helped them to go through it
step-by-step if needed. Then we discussed what kind of benefits this application might bring for the females
to fight their harassment on Facebook Messenger in different kinds of cases. We also asked them what kind of
challenges and troubles this application might generate for the users and other stakeholders.
The sessions were conducted in the Bengali language. The average length of the sessions was 35 mins. We
requested for participants’ permission to audio record the sessions and received permission to record 5 of them.
18
Unmochon
Details of Post-Design User-Evaluation
FGD
Total Number of FGD Participant: 29 (Female:19, Male:10)
Age Range (in Years) Occupation
18-25: 25 (86%) Public/private job-holder: 17 (59%)
26-35: 4 (14%) Student: 12 (41%)
36+: 0 (0%) Others: 0 (0%)
Interview
Total Number of Interview Participants: 19 (Female:13, Male:6)
Age Range (in Years) Occupation
18-25: 6 (32%) Student: 6 (32%)
26-35: 11 (58%) Public/private job-holder: 13 (68%)
36+: 2 (11%) Others: 0 (0%)
Table 2. Demographic Characteristics of Participants of Post-design User-Feedback
We also offered them to discard the discussion at any moment due to their discomfort. However, no such incident
happened.
5.1.2 Interviews. Along with the FGDs, we conducted 19 interviews (13 females and 6 males) with Facebook
users as they requested a one-on-one conversation instead of a group discussion. 8 of them were from the set of
participants of pre-design interviews. As before, we recruited the rest of the participants from our professional
networks on social media, from Facebook groups and friends. Upon reaching out to them, we explained the
purpose of the work. We also sent them the user-study package a few hours before the interview so that they
experience those and get some time to think about those. During the interviews, we explained to them the
purpose of the application and helped them to go through it step-by-step if needed. Then we asked them what
kind of benefits did they see and how this application might be effective for different kinds of gender harassment
scenarios on Facebook messenger. We also asked them what kind of challenges and troubles they could imagine
that might generate from this application for the users and other stakeholders.
All of the interview sessions used Bengali as the primary language. It generally took around 25 and 40 minutes
to complete the interview. 13 of the interviews were audio-recorded with the permission of the participants. We
left every opportunity for the participants to leave the interview if they felt uncomfortable, even during any
ongoing sessions. We also informed them that we would discard the record of their participation if they wanted.
However, no such event took place over the course of the study.
5.2 Data Collection and Analysis
The audio files of interviews were recorded in the researchers’ phone and later saved in a hard drive for further
data processing steps. We collected a total of approximately 8 hours of audio recordings and around 100 pages
of interview-notes. In the analysis process, we first transcribed the audio recordings and later translated them
into English. We then performed thematic analysis on the transcriptions and our detailed notes [15, 85]. Four
of the authors independently read through the transcripts carefully and allowed codes to develop. Later they
shared their codes with each other. A total of 24 codes spontaneously developed during the first round of the
coding. Then we clustered related codes into themes after a few iterations. Some of the themes seemed recurring,
for example, vindictive, privacy, posting, recognition, justice, etc. Such themes influence the organization of our
finding section presented next.
19
6 FINDINGS FROM USERS’ FEEDBACK
We organize the findings from the the user evaluation in three categories: design concerns and suggestions,
justice concerns, feedback on user interface and process. We discuss them below.
6.1 Design Concerns and Suggestions
One of the suggestions that came out during an FGD session with 6 participants was whether it is possible to check
the metadata and include a verification status with the screenshots before sending that to the server and adding
that in the reports’ status while posting them on Facebook. They explained that this approach could advantage
the authenticity of the evidence presented by the victim. In another group of 5 participants, they discussed what
could be the disadvantages of letting the harassment reporters hide their identifiers by drawing lines on them.
They discussed a hypothetical scenario where someone manipulated the browser with its meta-elements, then
captured the screenshot and hid the identifiers. In the same conversation, one of the participants suggested a
possible solution to such scenario could be to let the application report the screenshot along with metadata of
the page. Also, they suggested that the tool should reload the page by itself automatically before it captures
the screenshot, to avoid possible manipulation in the chat through some advanced engineering with the HTML
elements of the page. Concerning this point, one of them suggested developing this as a mobile app.
10 of our FGD participants anticipated that once such a tool is available, the server will be flooded with reports
and many of those might be just random people sending random reports which might have nothing to do with
harassment. They also proposed that this tool needs an intelligent filter to select and let in harassment-related
posts only. Also, in these conversations, the participants discussed the need for a guideline to define harassment.
Another group of 4 participants raised a concern that if the harasser is wise, then they would immediately know
which one of the victims of them have reported just by seeing the screenshot in the report. They also suggested
to report the name of the harasser upon verifying the authenticity of the evidence and not to expose the image of
the chat; to hide all the possible identifiers of the victim.
6.2 Justice Concerns
During the user evaluation, our participants also discussed and showed their concern about hegemonic legal
consciousness of the stakeholders in the conversation of gender justice; many aspects of mob justice on social
media and how those influence gender justice; and offenders’ privacy in this regard. We discuss them below.
6.2.1 Hegemonic Legal Consciousness. One of our interview participants, P61, is a lawyer and she informed
us that institutional justice in Bangladesh is not transparent and female-friendly. The system is inclined more
towards power-politics. Thus, often the victims find it easier to seek justice in social shaming. She told us some
of the stories from the cases of gender harassment which she recently handled and shared her experience with us
as a relevant example. A female debater developed an intimacy with another fellow male and that affinity turned
toxic at some point. Later, the female debater turned in a written complaint to their association and accused the
male colleague of gender harassment on messenger inbox in a particular social media. As P61 explained,
“The female submitted some screenshots of their chat history to their association to make the case of
harassment. Then the accused offender submitted some evidence from earlier chat histories to prove that
they were in a romantic relationship at some point and claimed that the female colleague is accusing him
of harassment to break his upcoming marriage. On the basis of the submitted evidence, the association
came to the conclusion that the female colleague was NOT harassed, but the male colleague should also
have behaved himself." (P61)
P61, thus, explained how gender, social position, and other hegemony influence the transparency in the
model of justice in Bangladeshi society, including professional spaces. She further stretched this conversation
20
Unmochon
and explained how such a social and professional setting often disregards women’s sensitivity to their online
harassment. She pointed out that it would be generally difficult for women to seek out online gender justice,
even with technology like Unmochon.
6.2.2 Mob Justice on Social Media. Our participants brought up several challenges while seeking gender justice
on social media. During our discussion with one of the male participants, he informed us that communication
skill matters the most in seeking online mob justice. He explained,
“When a female victim exposes the online harasser to the online community shared by both of them,
often her communication skill plays a role here. Sometimes people visit the victim’s Facebook profile and
thoroughly investigate if she looks suspicious. If she wears modern clothes, then people often redirect the
case and say she was asking for it. Often women lose control over their temper at that point and start
fighting. However, if the victim stays calm, ignores the slut-shamming, and responds to the audiences
with relevant evidence then the victim wins their support." (P73)
In the same discussion, he added that storytelling is also equally important. When a woman keeps telling a
consistent story on social media and if the clues that she leaves connect to reality and to the version spoken by
the other stakeholders, then that story is widely accepted and the victim is likely to be trusted by the people.
Therefore, how the evidence is compiling into a coherent narrative rather than a scattered truth often makes the
difference, as he explained,
“There could be multiple versions of the same story. But it is the woman’s burden to speak out the version
in a way that aligns with plausibility. If she sounds strange and illogical in view of the public, then her
claims are denied. For example, a group of words, “I really need to meet you this evening, if you do not
meet me then that is going to be a problem" could mean a lot of things. People will be curious about
why the accused person wanted to meet her. People on Facebook would say it is her job to provide more
context and coherent evidence so that it can be categorized as harassment and not some business or real
needs." (P73)
8 of the interview participants also were concerned by the standard of judgment by the people on Facebook as
there are diverse groups of people with different ideologies and philosophies. They were concerned about who
would decide the metrics in a complex social setting of multiple moralities, as they expressed it this way,
“There are different types of people on Facebook. Some are extremely patriarchal, some are extremely
conservative. There are also people who are against women using Facebook since they think women
meeting unknown people on Facebook, losing their purdah and this is against the women’s purity concept
in their practicing Muslim ideology. So when a woman comes and seeks justice here, if she has any such
people in the community in front of which she is trying to establish her requited gender justice, they
might face a backlash as the mob would say she should have not talked to men or create a scope to let the
men talk to her. They might say she invited it. So in such cases, even if she proves that the harassment
was true, she might still fail to establish justice because of this blame." (P64)
13 of our participants pointed to the fact that the idea of exposing the offender in front of the online community
shared by both of the parties using the screenshot of chat history repetitively might lessen the gravity of their
complaints and could be problematic for them. Thus, the community is involved in the process of investigation
and may become curious about further details which might be irrelevant to the complaint made, but later those
can be subject to some other harassment. For example, P67 explained the story of one of her female friends,
“A few months back, my friend first complained against a male classmate in our class’s Facebook group
with no solid proof. Many of the group-members found that baseless and dismissed her accusation.
Recently, when she complained against the same person again, the harasser and his fellows said that she
21
was deliberately doing this again and pointed to the previously dismissed complaint. Many people in the
group then dissed her recalling her previous complaint, my friend felt sick and left the group." (P67)
Thus, our participants informed us through many discussions that securing online gender justice with mob
would need not only authentic evidence but how the victim communicates and how they align with mob sentiment
are equally important factors in this regard. They further stretched the discussion and urged that our design
goals should include such needs of the victims for the upcoming versions of this prototype.
6.2.3 Offender’s Privacy. 5 of our interview participants also pointed that the harasser might make a case of
their privacy breaching if a victim share their personal conversation in front of a wider audience. One of the
female participants shared her experience of accusing a male friend of harassment and further consequences,
“First when I shared the photos of our chat history with some of the other friends, he claimed that the
photos are made up. Like I fabricated them. Then one day I logged into my Facebook profile in front of
people at the university and proved that I did not manipulate it, he started countering me by saying
whatever happened, he did it privately and yelled at me asking why did I invade his privacy in front
of people. Do I really have to care about his privacy after he had done so much damage to my mental
health?" (P59)
While making this point, the participant also added that the tool we have developed, may not be able to secure
the privacy of the harasser and might be much less helpful than expected by the victim while seeking gender
justice for their online harassment.
6.3 Feedback on User Interface and Process
During the user-evaluation, the participants gave us some feedback on the User Interface. For example, 5 FGD
participants showed their frustration on the lengthy processing time of the image after being captured by the
tool. Also, another 6 participants suggested having a progress bar so that the users may know the approximate
waiting time for particular processes. 9 of the FDG participants suggested to convert it into a mobile application
and put it as a floating button on top of the screen. They added that a lot of users in Bangladesh do not have
access to a desktop. Therefore, developing it as a mobile application would benefit them. 11 of the participants
insisted on making the application more informative, interactive, and guided. During the interview session with
P71, she explained her problem,
“While running the application, I did not understand if it had already taken the screenshot and where
did it go. I think it will be more helpful if the application windows are a bit more verbose and tell the
users what exactly is happening in the ongoing step." (P71)
The required browser version to run tool did not match 5 of our participants’ browser versions and thus, they
had a hard time using the tool. After testing for some more time, participant P60 noticed that while capturing
a screenshot, it’s difficult to move the application window from the intended surface for the screenshot. Both
participants P61 and P59 tried but the application window could not be moved or minimized once the button was
clicked.
7 DISCUSSION
In this paper, we have described the details of Bangladeshi female internet users’ gender harassment and what
kind of measures do they take while addressing those. Then drawing on our findings, we built a set of design
goals that are more practical and appropriate for helping the victims fight their online gender harassment. We
prototyped a version of Unmochon, conducted user-evaluation, and reported the perception of victims and other
stakeholders of such feminist designs. Findings of the work lead us to discuss some broader agendas of feminist
technology design, social support and law, and gender justice.
22
Unmochon
7.1 Feminist Technology Design for the Global South
Our study joins the feminist-HCI discussion in the context of the global south. We cohere with some previous
literature and show that women in this particular region are socially and culturally trained to accept or ignore
their harassment and not to seek out external help [3, 88]. However, women in such strongly patriarchal societies
still find ways to cope with the situation and fulfill their needs even within their adversarial environments [86, 87].
Some of the feminist-HCI literature in this social setting argue to design for women using all possible local support
available for the women and enable other forms of situated tactics [87, 88]. Our design of Unmochon adapts
the assumptions from ‘designing within patriarchy’ orientations [88] and joins this body of feminist technology
design [8]. Feminist-HCI calls for establishing a relationship with subjects [8]. Similarly, researchers working on
gender harassment on social media often call for engaging more women, especially gender harassment victims in
designing gender justice tools and techniques [12]. We have responded to these calls since the majority of the
authors and designers in this work are females and have experienced gender harassment on social media. We
believe that our perspectives have helped us understand the participants and integrated cultural appropriateness
in the features and affordances in designing tools to support gender justice. Building on the findings from our
study, we argue that along with focusing on the users and their immediate needs, feminist-HCI and ICTD should
also aim towards political agenda like supporting gender movements on social media for a better online experience
for the female social media users from such complex patriarchal societies.
7.2 Social Support and Law
Our study of online gender harassment also joins the design for social good movement within computer science,
Ubicomp, and HCI [1, 10, 19, 31, 34, 72]. Building on the concept of community-based social support, the existing
body of work adapts the assumptions of affirmative design and persuasive design [32, 44, 74] and contributes to
the goals of social good for all. With our study, we have extended this line of literature to the context of the global
south where often users and multiple stakeholders hold diverse and tangential agendas in a given social setting
[3]. In this paper, we show that although often women seeking gender justice online are dissed and humiliated by
online populations, they still seek support from the community via friends or using a fake Facebook ID. We have
also showed that victims often believe that their sporadically and locally organized small-scale protests might
help them more than legal systems and we have listed a number of reasons for their belief. All of these angles
rising from the findings of our study lead us to argue that technology should be designed aiming to balance and
bridge between social support and legal perspective to support the victims better.
7.3 Shaming Platform and Transformative Justice
Finally, we highlight the transformative aspect of the design that we implemented through ‘Unmochon’. Transfor-
mative design is based on a broad idea of fairness. Instead of punishing an individual for a crime, transformative
justice investigates why that person committed that crime. This allows us to find deeper problems in our society
including poverty, discrimination, lack of education, lack of social support, lack of goo childhood, etc., which
contribute to making a person criminal [23, 37, 67]. Transformative justice, hence, focuses on fixing on those
bigger issues [22]. This happens in three channels: (a) restoring the social imbalance caused by the individuals
through social support, (b) changing the mindset of individuals who committed a crime through education and
social support, and (c) addressing the bigger social problem that contributed to making a person commit crimes.
Thus, transformative justice provides a way to reduce crime in a society in a more sustainable way.
One of the design principles of Unmochon is ‘shaming’ which may be interpreted as an adversarial action
toward punishing someone for their misconduct. However, shaming also provides a person with a way to reflect
on what they have done from their social peers and repent for that. Therefore, shaming may reduce crime in a
society [16]. The social process of shaming often also involves seeking forgiveness and committing not to repeat
23
the crime. At the same time, social pressure is imposed on the person that keeps them away from a repetition of
the same crime. The perpetrator then gets an opportunity to learn and change their behavior and mindset. At the
same time, the discourse that is created around a shaming incident allows the society to reflect on the broader
picture of the society - where the perpetrators are coming from? why is this happening? how can we put a stop
to this? how can families, communities, and societies come together to fix the problem?. Such conversations
often lead to social welfare activities. From this perspective, we hope that Unmochon also creates a platform for
transformative justice. Having said that, we do also acknowledge that the actual happening of transformative
design largely depends on how people will use this platform and there are chances that many people may only
focus on the immediate punishment part of shaming. However, we believe that, given enough time, more and
more people will start focusing on the long-term sustainable solutions of the problem and the transformative
aspects of Unmochon will prevail.
8 LIMITATIONS OF THE WORK
Our work has several limitations. First, both online and real-life gender harassment is a stigma-topic in Bangladesh
and for that reason, recruiting participants was challenging for us.Most of the interview and focus group discussion
participants are recruited through snowball sampling and thus they are mostly the people from the researchers’
primary and secondary networks. Since we recruited the participants through convenient sampling, our work
is not free from participation bias and selection bias. The opinions of our participants and arguments driven
from those may not represent the collective view of the females of the whole nation. Second, we failed to engage
with any participant of the third gender due to our limited reach. Thus, we did not gain any insight into their
experiences of harassment and their opinion about our tool. Despite these limitations, the findings of our study
will be useful for feminist technology design in the context of low-resource and patriarchal setting. Also, the
arguments and lessons from this study will contribute to women-empowerment policy-making in Bangladesh
and other countries with a similar resource, environment, and social setting.
9 IMPLICATIONS FOR DESIGN AND FUTURE WORK
In the future iterations of the design, we plan to overcome or minimize some of the limitations of the existing
version. Our user-evaluation also points to some of the necessary future design iterations. First, we prototyped
our tool for Facebook messenger only. Our participants mentioned that they use Imo, Viber, Whatsapp, and
other text messaging services frequently and face gender harassment on those platforms as well. Following their
suggestions, one of the future work could be developing similar tools for Imo, Viber, Whatsapp, and other text
messaging services that they use. Second, Our participants pointed out that sometimes it is difficult to explain
the whole story of harassment just using one image and thus often the offenders go away taking the benefit
of the doubt. In the existing version of the prototype, there is no way to collage multiple screenshots in the
same report to explain the flow of the conversation. Our participants requested such functionality in the tool’s
future iteration of the design. Third, this version of the prototype does not support submitting the evidence to
police directly, although it fulfills some of the qualities and matches the protocol of the standard legal evidence
collection followed by the Bangladeshi police in cases on cyber-crimes. This opens up a new opportunity for us
to design technologies supporting gender justice within legal spheres. Finally, we will find ways to set up better
communication between the victim and the law-enforcement authorities through this application and revisit our
design assumptions in this regard.
10 CONCLUSIONS
In this paper, we present a deeper understanding of online gender harassment faced by Bangladeshi female social
media users. First, we conducted an extensive investigation of women’s online experiences of harassment. Building
24
Unmochon
on the findings from that, we then set our design goals to support gender justice on Facebook often sought by
female victims and designed Unmochon application for women who were harassed on Facebook messenger. Our
user-evaluation with female Facebook users and some other stakeholders suggests a further improvement in
design assumptions, goals, and the user interface. We further extend our discussion on a different aspect of legal,
ethical, and pragmatic trade-offs of this type of approach. Taken together, our findings expand the ICTD and
feminist-HCI communities’ understanding of gender-justice technology design within a resource-constrained
and patriarchal social setting of Bangladesh and other similar countries.
11 ACKNOWLEDGMENTS
Anonymized.
REFERENCES
[1] Rediet Abebe and Kira Goldner. 2018. Mechanism design for social good. AI Matters 4, 3 (2018), 27–34.
[2] Nova Ahmed, Sadd Azmeen Ur Rahman, Rahat Jahangir Rony, Tanvir Mushfique, and VikramMehta. 2016. Protibadinext: Sensor support
to handle sexual harassment. In Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing:
Adjunct. 918–921.
[3] Syed Ishtiaque Ahmed, Steven J Jackson, Nova Ahmed, Hasan Shahid Ferdous, Md Rashidujjaman Rifat, ASM Rizvi, Shamir Ahmed, and
Rifat Sabbir Mansur. 2014. Protibadi: A platform for fighting sexual harassment in urban Bangladesh. In Proceedings of the 32nd annual
ACM conference on Human factors in computing systems. ACM, 2695–2704.
[4] Saad Ahmed Akash, Md Al-Zihad, Tamal Adhikary, Md Abdur Razzaque, and Arifa Sharmin. 2016. Hearme: A smart mobile application
for mitigating women harassment. In 2016 IEEE International WIE Conference on Electrical and Computer Engineering (WIECON-ECE).
IEEE, 87–90.
[5] Mohammed Eunus Ali, Shabnam Basera Rishta, Lazima Ansari, TanzimaHashem, and Ahamad Imtiaz Khan. 2015. SafeStreet: empowering
women against street harassment using a privacy-aware location based application. In Proceedings of the Seventh International Conference
on Information and Communication Technologies and Development. ACM, 24.
[6] Nazanin Andalibi, Oliver L Haimson, Munmun De Choudhury, and Andrea Forte. 2016. Understanding social media disclosures of sexual
abuse through the lenses of support seeking and anonymity. In Proceedings of the 2016 CHI conference on human factors in computing
systems. 3906–3918.
[7] UNWomen – Asia-Pacific. 2020. Facts and Figures. https://asiapacific.unwomen.org/en/focus-areas/end-violence-against-women/evaw-
facts-and-figures.
[8] Shaowen Bardzell. 2010. Feminist HCI: taking stock and outlining an agenda for design. In Proceedings of the SIGCHI conference on
human factors in computing systems. 1301–1310.
[9] Jamie Bartlett, Richard Norrie, Sofia Patel, Rebekka Rumpel, and Simon Wibberley. 2014. Misogyny on twitter. Demos (2014), 1–18.
[10] Russell Beale, Nicola Bidwell, Eli Blevis, Stephen Brewster, Anxo Cereijo Roibas, Keith Cheverst, Andrew Deardon, Jussi Impio, Amit A
Navavati, Abigal Sellen, Yvonne Rogers, and Lucia Terrenghi. 2009. UBICOMP 2009 Workshop CFP: Globicomp - Taking Ubicomp
Beyond Developed Worlds from Julie Kientz on 2009-02-04 (www-multimodal@w3.org from February 2009). https://lists.w3.org/
Archives/Public/www-multimodal/2009Feb/0000.html.
[11] Reuben Binns, Michael Veale, Max Van Kleek, and Nigel Shadbolt. 2017. Like trainer, like bot? Inheritance of bias in algorithmic content
moderation. In International conference on social informatics. Springer, 405–415.
[12] Lindsay Blackwell, Jill Dimond, Sarita Schoenebeck, and Cliff Lampe. 2017. Classification and its consequences for online harassment:
Design insights from heartmob. Proceedings of the ACM on Human-Computer Interaction 1, CSCW (2017), 1–19.
[13] Lindsay Blackwell, Mark Handel, Sarah T Roberts, Amy Bruckman, and Kimberly Voll. 2018. Understanding" Bad Actors" Online. In
Extended Abstracts of the 2018 CHI Conference on Human Factors in Computing Systems. 1–7.
[14] Jan Blom, Divya Viswanathan, Mirjana Spasojevic, Janet Go, Karthik Acharya, and Robert Ahonius. 2010. Fear and the city: role of
mobile services in harnessing safety and security in urban use contexts. In Proceedings of the SIGCHI Conference on Human Factors in
Computing Systems. 1841–1850.
[15] Richard E Boyatzis. 1998. Transforming qualitative information: Thematic analysis and code development. sage.
[16] John Braithwaite. 2000. Shame and criminal justice. Canadian Journal of Criminology 42, 3 (2000), 281–298.
[17] Tarana Burke. 2006. me too. Movement. https://metoomvmt.org/.
[18] Tarana Burke. 2018. Why We Need to Acknowledge the True Founder of the #MeToo Movement. https://www.blackburncenter.org/
post/2018/02/06/why-we-need-to-acknowledge-the-true-founder-of-the-metoo-movement-tarana-burke.
25
[19] Daniela Busse, Eli Blevis, Richard Beckwith, Shaowen Bardzell, Phoebe Sengers, Bill Tomlinson, Lisa Nathan, and Samuel Mann. 2012.
Social sustainability: an HCI agenda. In CHI’12 Extended Abstracts on Human Factors in Computing Systems. 1151–1154.
[20] Eshwar Chandrasekharan, Mattia Samory, Anirudh Srinivasan, and Eric Gilbert. 2017. The bag of communities: Identifying abusive
behavior online with preexisting internet data. In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems.
3175–3187.
[21] Rhitu Chatterjee. 2018. A New Survey Finds 81 Percent Of Women Have Experienced Sexual Harassment. https://www.npr.org/sections/
thetwo-way/2018/02/21/587671849/a-new-survey-finds-eighty-percent-of-women-have-experienced-sexual-harassment.
[22] Donna Coker. 2002. Transformative justice: Anti-subordination processes in cases of domestic violence. (2002).
[23] Erin Daly. 2001. Transformative justice: Charting a path to reconciliation. Int’l Legal Persp. 12 (2001), 73.
[24] Priyashree Dasgupta. 2018. #MeToo In India: 75 Professors, 30 Institutes, What Happened To Raya Sarkar’s List Of Sexual Ha-
rassers? https://www.huffingtonpost.in/2018/10/25/metoo-in-india-75-professors-30-institutes-what-happened-to-raya-sarkar-s-list-
of-sexual-harassers_a_23571422/.
[25] Damayanti Datta, Shweta Punj, and Chinki Sinha. 2018. #MeToo hits home - Cover Story News. https://www.indiatoday.in/magazine/
cover-story/story/20181022-metoo-hits-home-1360419-2018-10-12.
[26] DhakaTribuneDeskReport. 2019. 70% of women facing cyber harassment are 15-25 years in age. https://www.dhakatribune.com/
bangladesh/dhaka/2019/09/24/70-of-women-facing-cyber-harassment-are-15-25-years-in-age.
[27] Jill P Dimond, Michaelanne Dye, Daphne LaRose, and Amy S Bruckman. 2013. Hollaback!: the role of storytelling online in a social
movement organization. In Proceedings of the 2013 conference on Computer supported cooperative work. ACM, 477–490.
[28] Carl DiSalvo. 2012. Adversarial design. The MIT Press.
[29] Hande Eslen-Ziya. 2013. Social media and Turkish feminism: New resources for social activism. Feminist Media Studies 13, 5 (2013),
860–870.
[30] Deepa Fadnis. 2017. Feminist activists protest tax on sanitary pads: attempts to normalize conversations about menstruation in India
using hashtag activism. Feminist Media Studies 17, 6 (2017), 1111–1114.
[31] Maria Angela Ferrario, Will Simm, Peter Newman, Stephen Forshaw, and Jon Whittle. 2014. Software engineering for’social good’:
integrating action research, participatory design, and agile development. In Companion Proceedings of the 36th International Conference
on Software Engineering. 520–523.
[32] Brian J Fogg. 2009. A behavior model for persuasive design. In Proceedings of the 4th international Conference on Persuasive Technology.
1–7.
[33] Jesse Fox and Wai Yen Tang. 2017. Women’s experiences with general and sexual harassment in online video games: Rumination,
organizational responsiveness, withdrawal, and coping strategies. New Media & Society 19, 8 (2017), 1290–1307.
[34] Sarah Fox, Mariam Asad, Katherine Lo, Jill P Dimond, Lynn S Dombrowski, and Shaowen Bardzell. 2016. Exploring social justice, design,
and HCI. In Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors in Computing Systems. 3293–3300.
[35] Radhika Gajjala. 2018. When an Indian whisper network went digital. Communication Culture & Critique 11, 3 (2018), 489–493.
[36] Björn Gambäck and Utpal Kumar Sikdar. 2017. Using convolutional neural networks to classify hate-speech. In Proceedings of the first
workshop on abusive language online. 85–90.
[37] Paul Gready and Simon Robins. 2014. From transitional to transformative justice: A new agenda for practice. International Journal of
Transitional Justice 8, 3 (2014), 339–361.
[38] Cristela Guerra. 2017. Where did ’Me Too’ come from? Activist Tarana Burke, long before hashtags. https:
//www.bostonglobe.com/lifestyle/2017/10/17/alyssa-milano-credits-activist-tarana-burke-with-founding-metoo-movement-
years-ago/o2Jv29v6ljObkKPTPB9KGP/story.html
[39] Margaret Hagan, Nan Zhang, and Joseph’Jofish’ Kaye. 2012. Safe mathare: a mobile system for women’s safe commutes in the slums. In
Proceedings of the 14th international conference on Human-computer interaction with mobile devices and services companion. 47–52.
[40] Randi Lee Harper. 2016. Putting out the Twitter trashfire. Art+ Marketing.(13 February 2016). Retrieved September 8 (2016), 2017.
[41] Jasmin Hasanović. 2015. Ochlocracy in the practices of civil society: a threat for democracy? Studia Juridica et Politica Jaurinenisis 2, 2
(2015), 56–66.
[42] N. Hassan, M.K. Mandal, M. Bhuiyan, A. Moitra, and S.I. Ahmed. 2019. Nonparticipation of bangladeshi women in #MeToo movement.
In ACM International Conference Proceeding Series. https://doi.org/10.1145/3287098.3287125
[43] Naeemul Hassan, Manash Kumar Mandal, Mansurul Bhuiyan, Aparna Moitra, and Syed Ishtiaque Ahmed. 2019. Nonparticipation of
Bangladeshi Women in# MeToo Movement. In Proceedings of the Tenth International Conference on Information and Communication
Technologies and Development. 1–5.
[44] Eric B Hekler, Jennifer C Taylor, Steven P Dow, Michèle Morris, Faren J Grant, Sayali S Phatak, Don Norman, mc schraefel, and Dana M
Lewis. 2019. Exploring, Defining, & Advancing Community-Driven Design for Social Impact. In Companion Publication of the 2019 on
Designing Interactive Systems Conference 2019 Companion. 373–376.
[45] Eleanor Tiplady Higgs. 2015. # JusticeforLiz: Power and Privilege in Digital Transnational Women’s Rights Activism. Feminist Media
Studies 15, 2 (2015), 344–347.
26
Unmochon
[46] Hossein Hosseini, Sreeram Kannan, Baosen Zhang, and Radha Poovendran. 2017. Deceiving google’s perspective api built for detecting
toxic comments. arXiv preprint arXiv:1702.08138 (2017).
[47] Shampa Iftakhar. 2020. # MeToo in Bangladesh: Can You Change? Journal of International Women’s Studies 21, 2 (2020), 126–142.
[48] Nadia Ilahi. 2009. Gendered contestations: An analysis of street harassment in Cairo and its implications for women’s access to public
spaces. Surfacing: An Interdisciplinary Journal for Gender in the Global South 2 (2009), 56–69.
[49] Justin Jager, Diane L Putnick, and Marc H Bornstein. 2017. II. More than just convenient: The scientific merits of homogeneous
convenience samples. Monographs of the Society for Research in Child Development 82, 2 (2017), 13–30.
[50] Syeda Gulshan Ferdous Jana. [n.d.]. 7. Bangladesh: Social media, extremism and freedom of expression. TRANSNATIONAL OTHERING
GLOBAL DIVERSITIES ([n. d.]), 103.
[51] Shagun Jhaver, Sucheta Ghoshal, Amy Bruckman, and Eric Gilbert. 2018. Online harassment and content moderation: The case of
blocklists. ACM Transactions on Computer-Human Interaction (TOCHI) 25, 2 (2018), 1–33.
[52] Natasha Kabir. 2018. Cyber Crime a New Form of Violence Against Women: From the Case Study of Bangladesh. Available at SSRN
3153467 (2018).
[53] Cristina Kadar and Irena Pletikosa Cvijikj. 2014. CityWatch: the personalized crime prevention assistant. In Proceedings of the 13th
International Conference on Mobile and Ubiquitous Multimedia. 260–261.
[54] Yoshiro Kamitake. 2007. From democracy to ochlocracy. Hitotsubashi journal of economics (2007), 83–93.
[55] Semanur Karaman. 2017. Women support each other in the face of harassment online, but policy reform is needed | LSE Women, Peace
and Security blog. https://blogs.lse.ac.uk/wps/2017/11/29/women-support-each-other-in-the-face-of-harassment-online-but-policy-
reform-is-needed/.
[56] Naveena Karusala and Neha Kumar. 2017. Women’s Safety in Public Spaces: Examining the Efficacy of Panic Buttons in New Delhi. In
Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems. ACM, 3340–3351.
[57] George Kennedy, Andrew McCollough, Edward Dixon, Alexei Bastidas, John Ryan, Chris Loo, and Saurav Sahay. 2017. Technology
solutions to combat online harassment. In Proceedings of the first workshop on abusive language online. 73–77.
[58] N. Kumar, N. Karusala, A. Ismail, M. Wong-Villacres, and A. Vishwanath. 2019. Engaging Feminist Solidarity for Comparative Research,
Design, and Practice. In Proceedings of the ACM on Human-Computer Interaction CSCW. ACM, 167.
[59] Amanda Lenhart, Michele Ybarra, Kathryn Zickuhr, and Myeshia Price-Feeney. 2016. Online harassment, digital abuse, and cyberstalking
in America. Data and Society Research Institute.
[60] C Mallapur and A Alphonso. 2018. #MeTooIndia: 54% rise in sexual harassment reported at workplaces between 2014-17. https:
//www.indiaspend.com/metooindia-54-rise-in-sexual-harassment-reported-at-workplaces-between-2014-17/
[61] Claudia Manzi, Sharon Coen, Camillo Regalia, Ana Maria Yévenes, Cristina Giuliani, and Vivian L Vignoles. 2018. Being in the Social: A
cross-cultural and cross-generational study on identity processes related to Facebook use. Computers in Human Behavior 80 (2018),
81–87.
[62] Robert Meyer and Michel Cukier. 2006. Assessing the attack threat due to IRC channels. In International Conference on Dependable
Systems and Networks (DSN’06). IEEE, 467–472.
[63] Alyssa Milano. 2017. How We Can Help Women Come Forward. Time (oct 2017).
[64] Kathy Miriam. 2012. Feminism, neoliberalism, and SlutWalk. Feminist Studies 38, 1 (2012), 262–266.
[65] Durba Mitra. 2012. Critical perspectives on SlutWalks in India. Feminist studies 38, 1 (2012), 254–261.
[66] Aparna Moitra, Naeemul Hassan, Manash Kumar Mandal, Mansurul Bhuiyan, and Syed Ishtiaque Ahmed. 2020. Understanding the
Challenges for Bangladeshi Women to Participate in# MeToo Movement. Proceedings of the ACM on Human-Computer Interaction 4,
GROUP (2020), 1–25.
[67] Ruth Morris. 2000. Stories of transformative justice. Canadian Scholars’ Press.
[68] Fayika Farhat Nova, MD Rashidujjaman Rifat, Pratyasha Saha, Syed Ishtiaque Ahmed, and Shion Guha. 2019. Online sexual harassment
over anonymous social media in Bangladesh. In Proceedings of the Tenth International Conference on Information and Communication
Technologies and Development. 1–12.
[69] Rogers TE Orock. 2014. Crime, in/security and mob justice: The micropolitics of sovereignty in Cameroon. Social Dynamics 40, 2 (2014),
408–428.
[70] Ji Ho Park and Pascale Fung. 2017. One-step and two-step classification for abusive language detection on twitter. arXiv preprint
arXiv:1706.01206 (2017).
[71] Sangkeun Park, Sujin Kwon, and Uichin Lee. 2018. CampusWatch: Exploring Communitysourced Patrolling with Pervasive Mobile
Technology. Proceedings of the ACM on Human-Computer Interaction 2, CSCW (2018), 1–25.
[72] Nimmi Rangaswamy and Nithya Sambasivan. 2011. Cutting Chai, Jugaad, and Here Pheri: towards UbiComp for a global community.
Personal and Ubiquitous Computing 15, 6 (2011), 553–564.
[73] Elissa M Redmiles, Jessica Bodford, and Lindsay Blackwell. 2019. “I just want to feel safe”: A Diary Study of Safety Perceptions on Social
Media. In Proceedings of the International AAAI Conference on Web and Social Media, Vol. 13. 405–416.
27
[74] Johan Redström. 2006. Persuasive design: Fringes and foundations. In International Conference on Persuasive Technology. Springer,
112–122.
[75] Rebecca S Robinson. 2016. Pink Hijab Day: Mediation of the Hijab as a Symbol of Protest. International Journal of Communication 10
(2016), 20.
[76] Jennifer D Rubin, Lindsay Blackwell, and Terri D Conley. 2020. Fragile Masculinity: Men, Gender, and Online Harassment. In Proceedings
of the 2020 CHI Conference on Human Factors in Computing Systems. 1–14.
[77] Christine Satchell and Marcus Foth. 2011. Welcome to the jungle: HCI after dark. In CHI’11 Extended Abstracts on Human Factors in
Computing Systems. 753–762.
[78] Vijay Shankar. 2015. Announcing Facebook lite. Facebook Newsroom (2015).
[79] Ditilekha Sharma. 2018. What Is Missing In the #MeToo Movement? Economic and Political Weekly 53, 49 (2018). https://www.epw.in/
engage/article/what-is-missing-metoo-movement-limitation-law-justice
[80] Tamara Shepherd, Alison Harvey, Tim Jordan, Sam Srauy, and Kate Miltner. 2015. Histories of hating. Social Media+ Society 1, 2 (2015),
2056305115603997.
[81] Emily Shugerman. 2017. Me Too: Why are women sharing stories of sexual assault and how did it start?
[82] Sharon G Smith, Xinjian Zhang, Kathleen C Basile, Melissa T Merrick, Jing Wang, Marcie-jo Kresnow, and Jieru Chen. 2018. The national
intimate partner and sexual violence survey: 2015 data brief–updated release. (2018).
[83] Navya R Sogi, Priya Chatterjee, U Nethra, and V Suma. 2018. SMARISA: a raspberry pi based smart ring for women safety using IoT. In
2018 International Conference on Inventive Research in Computing Applications (ICIRCA). IEEE, 451–454.
[84] WHOA Comparison Statistics. 2013. Working to Halt Online Abuse. http://www.haltabuse.org/resources/stats/index.shtml.
[85] Anselm Strauss and Juliet Corbin. 1990. Open coding. Basics of qualitative research: Grounded theory procedures and techniques 2, 1990
(1990), 101–121.
[86] Sharifa Sultana and Syed Ishtiaque Ahmed. 2019. Witchcraft and HCI: Morality, Modernity, and Postcolonial Computing in Rural
Bangladesh. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems. ACM, 356.
[87] Sharifa Sultana, Syed Ishtiaque Ahmed, and Susan R. Fussell. 2019. “Parar-daktar Understands My Problems Better": Disentangling the
Challenges to Designing Better Access to Healthcare in Rural Bangladesh. Proceedings of the ACM on Human-Computer Interaction 1,
CSCW (2019), 25.
[88] Sharifa Sultana, François Guimbretière, Phoebe Sengers, and Nicola Dell. 2018. Design within a Patriarchal Society: Opportunities and
Challenges in Designing for Rural Women in Bangladesh. In Proceedings of the 2018 CHI Conference on Human Factors in Computing
Systems. ACM, 536.
[89] Kalpana Viswanath and Ashish Basu. 2015. SafetiPin: an innovative mobile app to collect data on women’s safety in Indian cities. Gender
& Development 23, 1 (2015), 45–60.
[90] Jessica Vitak, Kalyani Chadha, Linda Steiner, and Zahra Ashktorab. 2017. Identifying women’s experiences with and strategies for
mitigating negative effects of online harassment. In Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work
and Social Computing. 1231–1245.
[91] Charlie Warzel. 2016. A Honeypot for Assholes: Inside Twitters 10-Year Failure to Stop Harassment. BuzzFeed News (2016).
[92] Charlie Warzel. 2017. Twitter is still dismissing harassment reports and frustrating victims. Buzzfeed.(17 July 2017). Retrieved September
8 (2017), 2017.
[93] Beverly M Weber. 2016. Kübra Gümüşay, Muslim digital feminism and the politics of visuality in Germany. Feminist Media Studies 16, 1
(2016), 101–116.
[94] Sherri Williams. 2015. Digital defense: Black feminists resist violence with hashtag activism. Feminist media studies 15, 2 (2015), 341–344.
[95] Sherri Williams. 2016. # SayHerName: using digital activism to document violence against black women. Feminist media studies 16, 5
(2016), 922–925.
[96] Yohanes Sigit Purnomo WP, Theresia Devi Indriasari, Kusworo Anindito, Yoshua Andrean, and Jaka Galih Prasetyo. 2019. CrimeID:
Towards Crime Prevention and Community Safety in Indonesia using Mobile and Web Technology. International Journal of Interactive
Mobile Technologies (iJIM) 13, 09 (2019), 52–65.
[97] Ellery Wulczyn, Nithum Thain, and Lucas Dixon. 2017. Ex machina: Personal attacks seen at scale. In Proceedings of the 26th International
Conference on World Wide Web. 1391–1399.
[98] Susan P Wyche, Sarita Yardi Schoenebeck, and Andrea Forte. 2013. “Facebook is a luxury" an exploratory study of social media use in
rural Kenya. In Proceedings of the 2013 conference on Computer supported cooperative work. 33–44.
[99] Chelsea Young. 2014. HarassMap: using crowdsourced data to map sexual harassment in Egypt. Technology Innovation Management
Review 4, 3 (2014), 7.
28