The ACM case study on "Malicious Inputs to Content Filters" (ACM, N.D.) examines the case of "Blocker Plus", a company implementing content filters to ensure compliance of educational institutions with the Children's Internet Protection Act (CIPA). The company implemented machine learning algorithms in an attempt to avoid the labour-intensive task of maintaining a list of keywords and websites to block. The implemented system was exploited, resulting in content not directly failing under the CIPA provisions to be blocked. The present text will explore the case study by comparing it with the ACM Code of Conduct (ACM, 2018), the BCS Code of Conduct (BCS, 2022) as well as the revised CIPA (American Library Association, 2010) under the assumption that both ACM and BCS codes are binding for the company.

 With 54% of children under 13 having seen sexual content online (30% having seen sexual content online at school) (Common Sense, 2022) and almost half of children having seen harmful content online (Weale, 2022), CIPA provides provisions for the protection of minors while using the internet. To this extent, the act attempts to regulate content that is taken as obscene, child pornography or content considered harmful to minors, such as nudity, or that "lacks serious literary, artistic, political, or scientific value" (American Library Association, 2010).

 The actions and omissions of the company resulted in content that is considered of "literary, artistic, political, or scientific value" being blocked and specific groups being discriminated against. Apart from the company's moral obligations, this also resulted in ethical violations based on the two aforementioned Codes of Conduct. Under Articles 1.2 and 1.4, ACM instructs members to "Avoid Harm" and "Be fair and take action not to discriminate". Having no access to scientific information about vaccinations (being one the top most searched terms in 2020), the global warming effect (being an increasingly searched term), and matters such as homosexual weddings being censored violates the two above provisions (Google Trends, 2023).

 Moreover, the company failed to adequately review the developed system, as instructed by article 2.4 of the ACM Code of Conduct, lacked appropriate system impact evaluation as per article 2.5 and demonstrated a lack of public information provision about the wrong filtering, possible mitigation measures and system limitations, as per Article 2.7. At the same time, the proven system exploitation showed a lack of appropriate risk analysis and technological measures, directly violating Article 2.9, which states, "In cases where misuse or harm are predictable or unavoidable, the best option may be to not implement the system.", especially when considering the refusal of the company to roll-back to the previous system used after identifying the severe issues of the newly implemented system. The fact that the implemented system directly impacted society and the education system, and considering the strong indications of lack of adequate pre-implementation review, resulted in the violation of Article 3.7 as the company failed to "Recognize and take special care of systems that become integrated into the infrastructure of society". All of the above led directly to a violation of Article 4.1. as the company failed to uphold several vital parts of the Code, severely impacting the education system.

 Similarly, when comparing the actions and omissions of the company with the BCS Code of Conduct, we can find several direct violations. Even though the actions of the company had to be aligned with quotation 1.a. (considering that the company was implementing a CIPA compliance system), the company showed disregard for environmental matters by allowing censorship of the term climate change and failed to uphold the public interest by discriminating based on sexual orientation and gender identity. Moreover, not following CIPA provisions showed a lack of understanding of quotation 2.d., by effectively failing to comply with the legislation, and failed to uphold the no-harm principle as per quotation 2.f. Additionally, it showed a lack of due care, as otherwise required by article 3.a. and failed to take responsibility for its omissions as per article 3.c. of the same.

 Finally, one should consider that the Codes of Conduct mentioned above, even though binding, are not necessarily strict legal documents, and thus, they may be open to interpretation. Similarly, CIPA can be open to interpretation when we consider both the lawmaker's intent and the sense of justice and present norms in a society. Having said that, even though in-home/private content filtering may not be able to be regulated, having in mind the freedom of speech and opinion and possible internet censorship (McCarthy, 2004), it is apparent that both the actions as well as the omissions of the company resulted in several forms of discrimination (in direct violation of article 1.4 of ACM Code of Conduct against discrimination), and resulted in actual internet censorship, by limiting content which is of "literary and scientific value" and thus being against the provisions of CIPA.

 To sum up, the company should have taken actions to avoid discriminatory filtering by utilizing subject-matter experts and adequate testing before releasing the machine learning models in a production environment (Articles 1.2, 1.4, 2.4, 2.5, 2.7 of ACM Code of Conduct and Citations 2.d., 2.f of BCS Code of Conduct), have measures in place to avoid misuse and exploitation of their system and underlying algorithms (Article 2.9 of ACM Code of Conduct) and have fall back plans in place. Moreover, the company showed complete and utter disregard after the issues had been identified, thus being in direct violation of articles 3.7 and 4.1 of the ACM Code of Conduct, quotations 3.a and 3.c of the BCS Code of Conduct and most notably with the provisions of CIPA.

Based on peers' responses

 The ACM case study on "Malicious Inputs to Content Filters" (ACM, N.D.) which examines the case of the company "Blocker Plus" and their product of web content filtering, has been analyzed in terms of technical, ethical and legal compliance. During the said analysis (Koilakos, 2023), several compliance issues have been found, some being legal, such as non-compliance with CIPA based on the law's intent, and others being ethical, based both on the ACM's Code of Conduct (ACM, N.D.) and the BCS' Code of Conduct (BCS, 2022).

 Apart from the company's violations, the analysis provided mitigation recommendations, such as utilizing subject-matter experts and testing adequately before the software's roll-out. In addition to that, leveraging the "power of the many" is another approach during which the company would make the algorithm and design ideas publicly available and transparent to the general audience (or to a limited audience - to ensure that their intellectual property is safeguarded), possibly allowing them to identify flaws early on (Kanakavelu, 2023). Another element that the company could have used is the implementation of a risk assessment, during which they would "consider the risks of harm or misuse before implementing the system" (Biswas, 2023).

 Overall, the company showed complete and utter disregard for the possible (and actual) misuse of their system, failed to balance the equities between automation and the public good and did not utilize any of the available design principles or mitigation measures available to them.

Abstract

 In conclusion, ethics shall be in the main picture of computing and not be considered a secondary duty. With technology continuing to evolve, ethical dilemmas are expected to become more complex, as can also be seen through the increased ethical concerns of every subsequent industrial revolution. Concerning data-related computing ethics and the involvement of people with specific needs in humanitarian contexts, addressing such challenges requires organisations and professionals to think outside of silos, and it needs a collective effort that involves ethicists, humanitarians, policymakers, computing and data professionals as well as the population of concern. It is only through dialogue, public consultation, strict oversight and enforcement, professionalism and commitment to upholding the ethical principles that we can ensure that the third and fourth industrial revolutions will not be overshadowed by harm to society and that humanity will thrive and enjoy the benefits of the innovation and prosperity that computing has to offer.

Guiding questions to support planning, revising, and refining of a literature review as provided by the University of Essex

Q: What is the focus and aim of your review? Who is your audience?
 A: Our literature review has the subject of "How can technology be used to raise awareness on the protection of refugees?".
 Starting from the target audience, our literature review targets a broad spectrum of target groups, mainly revolving around the general public. Other target groups include humanitarian organisations, non-governmental organisations and human rights advocates while it aims to reach policymakers. The audience, already familiar with our subject, shares the concern about adequately protecting refugees and upholding their rights, while they have a shared interest in raising awareness and driving change.
 As technological advancements continue changing and revolutionising various aspects of our lives, we must explore how these innovations can be used in intersection with advocating for refugee rights. The aim of the literature review is bi-fold: To identify and analyse the existing technological interventions that have been and are currently used for the cause mentioned above, and, to bring together tech professionals, humanitarian workers, and the public to collaborate on how technology can be at service of advocacy for refugee rights.


Q: Why is there a need for your review? Why is it significant?
 A: Technology's rapid evolution presents challenges and opportunities for addressing global humanitarian crises. Among these crises, the rights of refugees stand as one of the most urgent, with millions displaced in perilous situations. Public awareness of refugee rights remains somewhat limited in many layers of society. Given the evolution of technology, as mentioned above, and its potential benefits, the question of how technology can raise awareness for refugees' rights emerges.
 The proposed literature review is significant for various reasons, such as providing systematic pathways and discipline-related knowledge on the use of technology for advocacy purposes and understanding of current efforts and practices, looking into the emerging technologies which might not have been utilised for advocacy reasons yet, despite their potential, but also highlighting associated risks taking into consideration the sensitive nature of the population of concern, due to their legal and physical protection status.


Q: What is the context of the topic or issue? What perspective do you take? What framework do you use to synthesise the literature?
 A: The ever-increasing number of displaced people fleeing war and persecution is one of our time's most pressing humanitarian challenges. The context for our literature review is to provide clarity on the intersection of technological advancements and refugee advocacy.
 Our working assumption is that technology possesses the power to reshape negative narratives and raise awareness. For this reason, we approach the literature review as an interdisciplinary topic of humanitarian interventions and technological capability.
 We follow a hybrid approach between the traditional and systematic literature review frameworks. The workflow starts by categorising the main technological interventions (e.g. communication, virtual reality, AI, etc.) and further analysing relevant literature in terms of theory and real-world examples with a particular interest in advocacy usages. Through this workflow, we aim to achieve a complete understanding of the technological usage regarding advocacy causes and, furthermore, identify gaps, opportunities and risks.


Q: How did you locate and select sources for inclusion in the review?
 A: The process of locating and selecting resources for the literature review has been following both an academic approach and an applicable-practice approach. We begin by reviewing generic information from organisations dealing with refugee-related issues. At the same time, we search through Google Scholar for journals and other articles for scholarly literature related to technology, advocacy and human rights.
 Depending on the specific issue that we were researching (e.g. social media and advocacy, technology and ethics, etc.), we used a combination of keywords such as "social media", "advocacy" "refugee rights", "awareness", and "risks". Preference was given to the most recent articles and articles from academic sources. In parallel, we dived into information provided by the UN, INGOs and technology organisations related to advocacy. Organisations such as IOM, UNHCR and the UN have been of particular interest.
 The process above ensures that the list of sources is credible. Resources were further evaluated regarding relevance and contribution to the literature review subject. The followed methodology provided the complete picture of the studied matter by combining followed practice, academic research, criticality and applicability.


Q: How is your review structured?
 A: The structure of this review is planned from generic to specific. Starting from generic information, we indicate trends in technology usage to highlight its importance in everyday life, and we continue by displaying basic metrics of displacement and the connection of technology with inequalities. Moving to the more specific content, we connect technologies such as social media and VR with the advocacy and awareness-raising cause while discussing specific ethical and implementation concerns.


Q: What are the main findings in the literature on this topic?
 A: The literature on leveraging technology to promote awareness about refugee rights has revealed several findings. The digital era has significantly shifted the information provision and how people learn the news, with mainstream media being complemented, if not replaced, by social media platforms, allowing for more spherical and worldwide dissemination of information. Moreover, virtual reality (VR) and augmented reality (AR) can be and have been beneficial to the cause by providing firsthand experience of living within a refugee camp.
 At the same time, concerns about the technology's usage are critical. Ethical concerns, especially regarding the privacy and security of refugees, and other issues, such as audience desensitisation, have been at the forefront of the literature review.


Q: What are the main strengths and limitations of this literature?
 A: The literature review has been constructed based on the available public materials. This leads to several strengths but also specific limitations. Some of the most critical strengths of the literature review are the diverse perspectives presented due to the usage of several resources from different disciplines such as the academia, UN, INGOs, and the technology sector; the presentation of real-world applications, without limiting the review in theoretical concept; and the presentation from a forward-looking perspective without restriction to the past or the present. Regarding limitations, the forward-looking approach can be proven to be a double-edged sword in case the potential is overemphasised over current applications. Additionally, there is a lack of literature discussing the long-term impact, and finally, there is an underrepresentation of refugee-led literature.


Q: Are there any discrepancies in this literature?
 A: Due to the "generic to specific" approach we have followed throughout the composition of the literature review, and as with any other interdisciplinary field, the available literature presents discrepancies mainly revolving around ethical consideration and technology success rate for advocacy reasons. More specifically, a general observation is that literature tends to agree with the power of social media but also presents the lack of sustained commitment and regular supporters. Virtual reality (VR) and augmented reality (AR) are also fields where discrepancies are observed with focus on the benefit-risk ratio of AR and VR applications and viewer desensitisation.


Q: What conclusions do your draw from the review? What do you argue needs to be done as an outcome of the review?
 A: It is evident that technology offers powerful tools to swift the negative narratives and advocate for refugee rights. If correctly used, digital platforms and advanced technologies such as AR and VR can educate the public and give a path for more inclusive policymaking. This potential should be assessed regarding the population it serves, and ethical dilemmas and challenges should be addressed. Finally, a synergistic approach must be followed by bringing together humanitarians, policymakers, tech-industry experts and refugees to implement appropriate campaigns.

 The realm of computing, and especially the discipline of data science, has as a prerequisite the effective and proper usage of data collection methods as a pathway to understanding the studied matter at hand. Throughout this week's reading, there are several methods and techniques described which can be employed by researchers to achieve their research goals and support the research results. For this reason, we will compare the methods presented in the unit's readings and assess which are suitable for the proposed research/dissertation based on the research questions, aims and objectives, as well as based on the envisioned data required for the desired outcome.

 Our approach will consider the technical and theoretical subject of using queuing theory in the humanitarian context, along with its specific limitations. Starting with a generic overview, the research has a mix of exploratory and descriptive approaches, aiming to expand existing knowledge and procedures for how queues are utilized by directly applying findings in queuing environments through a simulation, as well as exploring underresearched or unresearched areas of the matter at hand (Fleetwood, 2021). Considering the vague data needs of the research (due to its nature), the ill-defined data sources, primarily due to the uncommon, hardly accessible and sometimes unsafe environment that we explore, and issues that might arise in access to data (Dudovskiy, 2022), a combination of data collection methodologies will need to be utilized.

 Among the various data collection methods presented in this week's reading material, and considering the research subject particularities, it is of particular interest the use of case studies to evaluate the research subject (Dawson, 2015; Fleetwood, 2021). Case studies will provide in-depth insight into various subject elements, such as the studied population or applied queuing in different environments. Also, other qualitative methods, such as one-to-one interviews with subject matter experts, will be needed. Field experts will be able to provide insights on followed practices as well as experience-led guidance (Dawson, 2015). Text contextual analysis on the available related literature, especially considering the diverse studied environment in terms of culture, ethnicity, geography, religion and sex, will provide a better understanding of the challenges of the research, as well as specific areas that the researcher should be prudent of (QuestionPro, 2021).

 Additionally, despite data source and availability challenges, the research has as one of its objectives to implement an algorithm/program that can be utilized to plan queuing activities. Considering the experimental element of the selected research method, the developed algorithm can be used as an isolated simulation environment to provide analytics on which algorithm works best and will thus help in deciding the final deliverable. For this reason, primary data collection will be used in terms of algorithmic success (Dudovskiy, 2022). Secondary data collection to compare the algorithm's success with the actual queuing times will depend on data availability.

 In conclusion, the chosen blend of data collection methods and research approaches vastly depends on the objectives, aim and constraints of the proposed research. While the methods and approaches mentioned above are the most appropriate at this preliminary phase of the proposal and can provide a well-rounded understanding of the topic, the selected options might change based on the steering of the research and research-specific challenges.

Preparation/introduction of the literature review

 The world is currently facing the most significant displacement in history, with nearly 110 million forcibly displaced people, out of which more than 45 million are refugees, asylum seekers and others in need of international protection (UNHCR, 2022). At the same time, anti-refugee rhetoric and negative perceptions towards refugees tend to be more and more frequent, especially among the host-country population (Findor et al., 2021) , although research argues that those opinions do not monopolize the refugee discussions (Loescher, 2021). In an effort to shift this narrative, technology can be leveraged, primarily through the mainstream media (having the lion's share in the public's information) (McCann et al., 2023).

 As previously mentioned, some of the louder (but not necessarily numerous) public voices lobby for stricter border policies and immigration laws. In order to shift the scale, the use of existing technologies, coupled with technological innovation, can be used. Starting from mainstream media, the era of only-written articles is long gone, with images and videos slowly replacing writing (Silverman, 2019). On the other hand, creative storytelling, such as the article "A Game of Shark and Minnow" (Himmelman, 2013), is visually appealing to the audience and can be similarly used by the mainstream media to tell the stories of refugee journeys and relevant risks.

 Journalistic tools continue beyond creative storytelling, with augmented reality being previously used for humanitarian purposes; going back to 2004 in a United Nations documentary (Irom, 2018), one can find several use cases, for example, showing a refugee's journey from Syria to Greece. Finally, one of the most common technologies used nowadays, social media, having universal coverage, can further assist in shifting anti-refugee narratives, especially if used by influential people or well-known and respected organizations (Kent, 2018).

 On the other hand, the use of technology needs to be thoroughly assessed before being used. Technologies such as augmented reality may put at risk already vulnerable populations. For example, cases in which augmented reality is used to sensitize the general public on the refugees' living conditions may pose risks related to data protection (Jerome & Greenberg, 2021), such as the refugees' whereabouts, negative coping mechanisms used, or even revealing information about their specific legal and physical protection needs (e.g. GBV cases). Another risk may be the audience's desensitization due to overexposure, similar to the desensitization of viewers due to overexposure to violence (Krahe et al., 2015).

 To conclude, technology presents immense opportunities for raising awareness about the protection needs of refugees, asylum seekers and people needing international protection. Leveraging technology use cases such as augmented reality can help provide the general public with hands-on experience in the life of a refugee. Consequently, sensitizing the public can assist in applying pressure on global leaders to take more robust protection measures for the people in need. All of the above should be used by having the "do not harm" principle (UNDP, 2023) in mind and ensuring that the safety of the people is not jeopardized at any step of the process.

Abstract

 Despite the exponential advancements in technology, which drive the "economic growth" (Driskell, 2022) and help people stay connected and in spite of the social media penetration mentioned previously, social inequalities are on the rise, suggesting that the world is moving forward in terms of technological application metrics but backwards in terms of inequality metrics. Private wealth is rising, while public wealth has been on a decline since 1970. Income inequalities are found to be on levels similar to those found in the early 20th century, while gender inequalities in pay are only 4% lower than in 1990 (World Inequality Report, 2022). Other examples of untackled inequalities include the lack of access to healthcare for women and children (United Nations, 2023), while the Gini coefficient, which measures inequality, increased from 27% in 2021 to 28% in 2022, signifying a general increase in inequalities in Ireland (Ireland Central Statistics Office, 2023). The United Nations (2023) also reports that 20% of people have been discriminated against at least once in their lifetime. The social injustices, despite the progress of humankind, also disproportionally affect refugees, asylum seekers and people in need of international protection.

 Researchers constantly face complex and cross-discipline ethical dilemmas while conducting their research duties (National Research Council et al., 2005). Those challenges are becoming even more complex in the case of paid research, where intellectual property, civil and criminal liabilities, and responsibility towards the general public come into play. The intellectual property (IP) aspect depends on multiple factors, such as legal agreements, the researcher's funding, and national legislation (The University of Melbourne, N.D.).

 In our case study (University of Essex, 2023), we discuss the scenario of Abi, a researcher who receives funding from Whizzz to evaluate the nutritious value of their cereal. During the research, Abi realised that the data suggest that the nutritious value of the product is lesser than expected, and further findings suggest that the product may even be harmful. There, Abi faces several ethical dilemmas related to using more favourable data analysis and how his research will be used, even if Abi shows both the positive and negative aspects of his research. The case study is analysed assuming the ordering party holds the IP rights, although the European Commission suggest co-authorship (EC, 2005).

 Undoubtedly, Abi, as a researcher, holds the primary responsibility, regardless of the ordering party's research objective, to conduct his research duties ethically and legally. Having said that, the research methodology must be defined beforehand (Kelley et al., 2003) and not change throughout the process to favour any outcome. At the same time, Abi can and should conduct his research unbiasedly, so he should favour neither the positive nor the negative findings (University of Cambridge, N.D.). This means researching the full spectrum of the product's nutritious values and writing his report by fully disclosing all the relevant findings.

 Moreover, despite Abi not holding the IR of his work, he may be professionally and scientifically responsible for how his research will be used based on the signed agreement between him and his funder, his contractual obligations, the relevant institutional code of conduct and applicable legislation (EC, 2005; Valkenburg et al., 2020). The much-needed peer review, primarily due to the implications of the research to the broader public and the ethical implications, and professional audits shall also have a say in the research outcome (Kelly, 2014; EC, 2005).

 Another aspect is the researcher's responsibility towards the public regardless of who holds the IP rights. For example, the French Centre of Scientific Research (2016) instructs researchers that they are ethically obliged to make their complete research available to other researchers and the general public, while the European Charter for Researchers (2005) suggests that the complete research should be made available to the general public and the ordering party as applicable and adequately exploited. This needs to be further assessed due to complex underlying legal implications.

 In summary, regardless of the ordering party's objectives, the searcher holds the primary responsibility to conduct his duties ethically and legally. This means the full findings shall be disclosed, whether favourable or not for the ordering entity. No data alteration or misrepresentation shall occur. The researcher can also employ the available tools to seek advice (e.g. through an appropriate ethics committee or a professional body's ethicist) or even refuse to accept the assignment on the grounds of upholding ethical standards. Finally, the proper exploitation of the results falls under the researcher's responsibilities (EC, 2005).

Based on peers' responses

 Researchers frequently face complex ethical challenges, especially in paid research, including but not limited to legal liabilities, public responsibility, and intellectual property (IP) rights. These challenges are influenced by various factors such as funding sources and legal agreements, national laws, and professional practices and ethics enforced by professional bodies (National Research Council et al., 2005; The University of Melbourne, N.D.).

 A case study from the University of Essex (2023) presents such challenges and ethical dilemmas by introducing Abi, a researcher funded by Whizzz to assess the nutritional value of their cereal. Abi's research results indicate that the product's nutritional value is below the expected, and some ingredients are potentially harmful. The ethical and legal dilemma revolves around data representation, findings communication and its subsequent usage.

 First, regardless of the funder's intentions, Abi's primary responsibility as a researcher is to execute his research ethically, without bias and prejudice and present both positive and negative results. The research methodology should be predetermined (Kelley et al., 2003), ensuring no shifts in approach to favour any desired outcomes (University of Cambridge, N.D.). Although Abi might not possess the IP rights, and despite the fact that the European Commission (2005) suggest co-authorship of researches, he remains professionally accountable for the research's application. Institutional guidelines further influence this accountability, along with the researcher's legal agreement with the funder and relevant laws (EC, 2005; Valkenburg et al., 2020). Peer reviews and professional audits also significantly impact research outcomes (Kelly, 2014; EC, 2005).

 Furthermore, researchers must share their findings with the public, irrespective of IP ownership. For instance, the French Centre of Scientific Research (2016) states that researchers must provide their research to peers and the public. Similarly, the European Charter for Researchers (2005) recommends sharing the complete research with the public and relevant stakeholders.  Finally, as a professional practice, Abi should take into consideration elements related to data protection, especially for data received by the funder or produced during the research that qualify as personal or sensitive data, and, at the same time, ensure that the primary data, which they are the cornerstone of the analysis, are verified in terms of accuracy and reliability (Jabbal, 2023).  In a nutshell, irrespective of a funder's goals, researchers must ethically and legally uphold their responsibilities. This means fully disclosing all results without bias or alteration. In case of doubt, they can seek ethical advice or decline projects to uphold these standards. Properly using and sharing research results is also of paramount importance to ensure that the public interest is satisfied (EC, 2005).

Abstract

 Following a similar approach, our research proposal aims to understand how queues work in other disciplines, how computing can provide value in the utilisation and optimisation of queues, and how queues can be used in humanitarian contexts in complex environments for registration and assistance delivery activities. By answering this subset of questions, we aim to showcase how the widely used concept of queues can use the computing and data analysis disciplines to assist the humanitarian field with the problem of reducing waiting times for populations in perilous situations and apply proper prioritisation by using the technology and computing concepts in a reproducible and usable manner. This technological intervention can potentially assist more than 108 million displaced people worldwide (UNHCR, 2023).

Abstract

 Hypothesis Testing and Summary Measures & Chart Worksheets and Analysis (View embedded)

Abstract

 Throughout this module, I explored and familiarised myself with different research methodologies, ethical considerations in computing, and academic professional practices. Due to limited pre-existing knowledge, I faced challenges in understanding research concepts. Succeeding units introduced hands-on literature review and research proposal presentation, data collection methods, and statistical approaches, which improved my understanding. The final units, emphasizing research writing and project management, shaped my future plans. Throughout the module, collaboration with peers has also been beneficial, while logistical challenges mainly related to conflicting priorities hindered my performance. Overall, this module improved my research capabilities and helped me find a passion for paper writing, which I aim to pursue in the future.