WebSci '20: 12th ACM Conference on Web Science Companion

It is our great pleasure to welcome you to the virtual workshops associated with the 12th ACM Conference on Web Science (WebSci’20). This year, the conference theme is “Making the Web Human-Centric? New Directions in the Web and AI.” As the workshop chairs, we selected workshops that exemplify this theme to complement the conference contributions. The theme of making the Web more human-centric is especially fitting this year as the workshops are held in the worldwide COVID-19 pandemic backdrop making the Web an ever more human-centric medium for work, play, communication, and many other societal necessities. Furthermore, the COVID-19 pandemic and the racial injustice riots that erupted in some parts of the world have forced Web Scientists to examine new directions for the Web and AI. These workshops span topics ranging from inequality, AI on the Web for social good, personalisation and community, explanations for AI, cybercrime and cybersecurity, data immortality, and evolutionary thinking for the Web.

From its inception, the World Wide Web has been intended to be a force for social good. But there are many barriers and obstacles, a situation commonly characterized as the Digital Divide. Several billion people especially in the Global South do not have access to Internet/Web for reasons of lacking (affordable) infrastructure, poverty, low literacy, lack of digital skills, language, etc., and are thus digitally excluded. Also in the Global North, despite being technologically “advanced”, we see severe digital inequalities and power disparities, in part for the same reasons and in part due to the Web being exploited as a centralized surveillance and money-making machine, controlled by big parties such as states and big (tech) corporations, thus creating further inequalities and exclusion. This paper summarizes the main themes and insights from ongoing research presented and discussed at the WebSci'20 Workshop on Digital (In)Equality, Digital Inclusion, Digital Humanism.

The African Sahel has become a stage for increasing violent conflict since 2009. It is probably no co-incidence that the upsurge of these conflicts went together with the increase in digital connectivity in the region. In this paper I explore the relationship between radicalization, conflict and access to ICTs and social media in the Sahel. First I sketch how the social landscape has changed due to the advancement of the ‘digital highway’. We will see how the so-called hybrid organization, the democratization forces of the digital have influenced the formation of trans-national communities and through these influenced conflict dynamics. The second part of the paper explores the increase in violence and the potential role of the digital networks in the upsurge of this violence. Radicalization in different forms (ethnic, religious, national) is an important effect of the information flows that travel the digital networks and that seems to influence the increasing (ugly) violence.

As we shift paradigms in the relationship between digital technologies and international development, many issues will need to be rethought. In this paper, I look at the changing nature of the relationship between digital and inequality in the global South; in particular tracing the re-scoping from concerns just about the digital divide and exclusion, to broader concerns about digital justice that also cover adverse incorporation into digitalised development systems across economic, political and social spheres.

The world is rapidly changing — due to many inequalities in our neoliberal society, but also due to the current Covid crisis, to be followed by a severe economical crisis. These crises will enlarge the already existing inequalities all over the world. In particular the inequalities between rich and poor, between literate and illiterate people, and between North and South. This crisis is the cause of great stress.

In 2006 a new type of company was established, called the B Corp. This new legislation started in the US (currently in 40 States) and protected the company from the existing shareholder-driven mentality. Instead of working for shareholders only, companies were going to work for all of its stakeholders. The legal form for this was the Benefit Corporation, but existing companies could also select to perform an impact analysis, called the BIA (Benefit Impact Analysis). With this analysis companies could qualify to become a certified B Corp.

After 14 years since its inception, there are now around 3,000 B Corps worldwide and around 5,000 Benefit Corporations, mainly in the US. Examples of well-known B Corps are Patagonia, Ben & Jerrys, Danone, Triodos Bank and African Clean Energy (ACE).

Besides this development of a new kind of company that uses the slogan “Business as a force for good” there is also another trend going on in the world, having to do with the exponential development of technology. The speed in which technology develops surpasses the adaptability of the human brain to every new step, also causing stress in the world.

Knowing that technology can be used for good and for bad, and that both directions cannot very well be influenced or steered by governments, we can only hope that technological developments “for good” are faster and more numerous than the bad ones.

Now, the Business as a force for good and the Technology as a force for good seem to come together and could potentially support each other. If we can define profitable business “for good” and that business is a technology business helping to solve the above-mentioned inequalities, we are making a big step in the direction of a better world.

This keynote describes these two developments in more detail, trying to come to a definition of a profitable business case using technology (especially AI) to improve the lives of many.

This paper applies a trans-disciplinary analysis on the issue of data sovereignty, from an African perspective. The paper interrogates the residence of data and the African prerogatives for its processing. Harvesting from experiences in Zimbabwean health systems, this paper suggests that African governments can steward the collection and appropriate use of data resources, applying the principles of data sovereignty.

A majority of the world remain unconnected to the World Wide Web due to issues like low literacy and relevant information. This study presents Mr. Meteo, a system that provides weather information via voice calls in local languages to rural farmers in Ghana. The study used an interdisciplinary approach to identify relevant informational needs and socio-economic implications, and early end-user and stakeholder involvement. Mr. Meteo was deployed in Bolgatanga, Ghana and represents a novel design in terms of actual web data access to rural areas. The positive feedback from farmers, and stakeholder’s interest in continuity, shows this approach to be an appropriate method of development and implementation of information systems for rural areas; successful due to end-user and stakeholder involvement, focus on existing technologies, the use of voice technologies to mitigate the problem of illiteracy, and information relevance to end-users. This paper presents the methodology and results of this novel, practical, local-context ICT4D project,that has produced a viable information system for rural communities.

Digital welfare does not operate in a vacuum, but rather transforms non-digital and unofficial spaces of welfare provision. The digitalisation of welfare occludes the complex reality of poverty and erects digital barriers to accessing welfare. Digitalised welfare has not abolished face-to-face support, but has relocated it to unofficial sites of welfare. The apparent efficiency and simplicity of the state’s digital welfare apparatus, therefore, is produced not by reducing the ‘messiness’ of welfare, but by rendering it invisible within the digital framework. In this paper we compare two approaches to welfare digitalisation and identify three considerations for welfare service design that might reduce the digital barriers, re-build a sense of self-efficacy and increase service accessibility and inclusion.

Digital divide remains a major concern among indigenous people despite the rapid increase in mobile phone penetration. Beyond the appalling participatory gaps, the devaluing of humanity in the digital era is seen as a much bigger threat. This paper addresses the emerging challenges that are putting indigenous communities at a great disadvantage. The use of emoji in social media and its prominence as a visual language for delivering emotional expressions, gestures and action in a speedy manner is fast becoming a way of life. The universality of emoji poses a major problem to the sustainability of the indigenous cultural values as it imposes western cultural and social behaviour hegemony on the indigenous society. Interactions with remote rural communities in Borneo over the past twenty years have revealed the devastating effect that such an outside-in communication medium has on their cultural resilience. Diversifying the emoji design in adopting a local cultural flavour alone, as described in current literature, will not address the issue effectively. Without considering the whole socio-technical system at a macro-level, the dangers relating to the aggravated disconnect to their rich traditional way of life cannot be averted. We posit a holistic socio-technical systems innovation approach with the participatory involvement of the indigenous community. In this paper, we demonstrate the modeling of complex systems that embed the socio-cultural context without compromising cultural values.

This paper – as a part of a broader research on the policies and practices of digital literacy in relation to inclusion-, seeks to identify the central tensions in teachers´ practices related to the use of digital technologies in education. In Argentina, these tensions have increased, the moment when the pandemic compelled the educational system at all levels, to switch from the physical to the virtual classroom. Despite the recent date of these events, in this study we aim to reflect on the first findings. We will continue this research through interviews and reports analysis, in the coming period.

Education and Artificial Intelligence (AI) can be complimentary to each other. Over the period the state of education in India has been influenced by some good reforms in education system and its implications are immensely towards positive side. AI being the latest technological advancement can be an approach for sustainable, smooth and transparent solutions. AI enabled technology can fill the existing gaps in the present education system. This paper traverses the key issues of the Indian education system with the objective of proposing some solutions which are inspired by the AI innovations having the sustainability as a significant part of it.

With modern supply chains spanning the globe, materials or components that companies use in their products may be sourced from areas prone to injustice and human rights abuse. A major challenge stakeholders face is the gathering of accurate data regarding producers in these areas. In this paper we introduce CARPA, a web application designed to gather reports on incidents and initiatives related to responsible production through crowd-sourcing. We describe its user-centric iterative process of development as well as its design and how this is influenced by the application context. Finally we discuss the challenges faced and the way forward.

Using power to achieve political goals is not a new strategy. This paper unpacks observations about the use of power in the development of information and communication technologies (ICTs) in Africa. This analysis shows the use of unequal powers by non-African academics, development actors and technical experts (including the power to set the agenda, fund and build), embedded deeply in the current structures of ICT development for Africa. It also looks at how benefits accrue to non-local development actors and outlines some of the unmitigated risks for Africa.

Although literacy rates around the world have increased and there is an expectation that individuals who access web pages will be able to read their content, this is not always the case. The barriers that may be faced can be linked to the way the system is designed and content is written. There may be complex language or a layout that is dense, cluttered and lacks clear markers regarding the key points being made.

Many organizations have provided guidance for web developers and authors offering suitable ways to ensure those accessing a website or service will have a pleasurable experience. However, it appears that there are still websites hosting pages with dense text, convoluted instructions and little support for those with low levels of literacy. When considering poor reading skills, the cause may be due to many factors including a lack of education, sensory and /or intellectual impairments and specific difficulties such as dyslexia. This means that the vast majority of online content may be hard to understand for a significant proportion of the world’s population. Moreover, these individuals may also lack digital skills, with little realization that assistive technologies and the availability of supportive access strategies can be helpful in these situations.

This paper aims to introduce the idea of enhancing readability of web content by using artificial intelligence (AI) techniques, such as linked data, natural language processing and image recognition to make available a wide range of automatically mapped multilingual symbols that can be used to clarify text content. In the past only a few symbol sets have been mapped and it was not possible to consider their appropriateness for text to symbol translations in a wide range of languages and cultural settings.

In this paper we discuss how to improve business sustainability of services for digital inclusion through value modeling and analysis using the e3-value method. Two questions come up: is this method understandable and useful in practice for ICT4D practitioners and developers, and is this method instrumental for development of sustainable services for digital inclusion? To answer this, three ICT4D student projects were carried out, that aim to improve digital inclusion in communities in Sarawak, Malaysia. Results show that the e3-value method is easy to learn and use in practice. It is instrumental (i) for visual conceptualization, facilitating discussion and co-construction of different business scenarios; (ii) it allows to assess potential profitability in the value web; (iii) for optimization of the system design (iv) to analyse strengths and weaknesses in the value network in terms of digital inclusion.

This year the AI3SD Network+ (Artificial Intelligence and Augmented Intelligence for Automated Investigations for Scientific Discovery) will be running a workshop at the WebSci ’20 Conference in Southampton, UK. Artificial and Augmented Intelligence systems have the potential to make a real difference in the scientific discovery domain however this brings a new wealth of ethical and societal implications to consider with regards to this research (e.g. human enhancement, algorithmic biases, risk of detriment). This workshop looks to explore the ethical and societal issues centered around using intelligent technologies (Artificial Intelligence, Augmented Intelligence, Machine Learning, and in general Semantic Web Knowledge Technologies) to further scientific discovery, with a strong consideration of data ethics and algorithmic accountability. Advances in technology and software are rarely inherently bad in themselves, however that unfortunately does not preclude them from being subverted to ill intent by others; furthermore, as demonstrated by the examples above, even an unintentional lack of care towards ethical codes and algorithmic accountability can lead to societal and ethical implications of scientific discovery. It is our responsibility as researchers to consider these issues in our research; are we conducting studies ethically? What ethical codes can we put in place for scientific discovery research to mitigate against ethical and societal issues. These are really important issues, and they require an interdisciplinary focus between scientists, social scientists and technical experts in order to be comprehensively addressed. AI4Good is a day long workshop including five keynotes, discussion sessions and an interactive activity. The first keynote is from Dr Will McNeill, from the University of Southampton. Will is a lecturer in Philosophy, and he will speak about Ethical Frameworks and Ethical Judgements. The second keynote will be given by Dr Cian O’Donnovan, a Researcher at UCL. Cian’s research is based on understanding how the benefits of emerging technologies can best contribute to a flourishing world. Cian’s talk will be on AI Ethics from the Ground Up: Cultivating Capabilities for Care. The third keynote will be given by Jacqui Ayling, a PhD Student at the University of Southampton. Jacqui’s PhD is on the topic of researching data protection and innovation in smart cities. She will be talking about Data Ethics for AI & Algorithmic Accountability. The fourth keynote will be given by Dr Peter Craigon from the University of Nottingham. Peter is a Research Fellow specialising in Ethics, and his talk will focus on the Moral IT Cards that he has developed. The final keynote will be given by Dr Samantha Kanza, an Enterprise Fellow at the University of Southampton. Samantha coordinates the AI for Scientific Discovery Network and developed a keen interest in the ethical and societal issues of technology whilst completing her PhD in Web Science. She will be presenting on Ethics for AI for Scientific Discovery.

To build a human-centric Web, we need a solid understanding of human connections online and of mechanisms for fostering such connections. People both perceive and respond to digital technologies and form connections with one another in diverse ways. Whether the goal is to improve engagement or to foster community, modelling user groups to personalise or tailor experiences can be key. Tailoring and connection-building are particularly crucial in spaces such as healthcare and education, where evidence clearly shows that perceived social support can facilitate learning and enhance outcomes.

PC’20 is a half-day workshop including two invited speakers, two discussion papers, and a broader discussion session.

The first keynote is from Su White, an Associate Professor at the University of Southampton. She will speak about the facilitation of education online.

The second keynote is from Michael Fergusson, CEO of Ayogo Health Inc. He will speak about “The Architecture of Choice: Using Psychosocial Variables to Dynamically Tailor Interventions.”

A session will highlight two discussion papers. Firstly, Jennifer Golbeck from the University of Maryland will present her work “Improving Emotional Well-Being on Social Media with Collaborative Filtering”. Secondly, Roushdat Elaheebocus, Poovanen Seenan, Sheekah Beharry and Girishsing Caussyram from the University of Mauritius will present their work “BehaviourCoach: A Customisable and Socially-Enhanced Exergaming Application Development Framework”.

Finally, a broader discussion session will consider: tools, techniques and case studies in tailoring and community-building; how these relate to one another; and next steps for Web Science researchers.

The workshop builds on health and education communities established through previous Web Science conference workshops. By using these two domains to ground discussion of user modelling and community, we intend to reinvigorate these communities.

A summary of the workshop will be created and shared online within two weeks of the event.

We thank the members of our program committee: Stéphane Bazan (TomKeen) and Charlie Hargood (University of Bournemouth).

Recommender systems and content personalization systems use algorithms that optimize for a given factor.We are interested in developing content recommendation algorithms that optimize for user well-being. In previous work, we showed that content can have large, significant impacts on users’ well-being. In this paper, we present the results of two large studies that show (1) people sometimes come to social media with the specific goal of improving their well-being and (2) that personalization systems can effectively recommend social media content that improves well-being.

The lack of physical exercising activities and improper nutrition are known to be among the leading factors which cause health issues. Mauritius, though a small island, ranks among the first in terms of affected ratio for these health conditions. This research looks into the development of a customisable application to promote sustainable behaviour changes in terms of physical exercises and nutritional habits through the use of gaming components alongside social and community-based elements. Evaluation of prototypes have shown promising results in terms of technical feasibility and users’ interests.

Individual healthcare choices are strongly influenced by psychosocial factors such as values, goals, and self-perceptions. By using validated instruments to measure and map psychosocial variables, it is possible to produce digital interventions that are dynamically tailored to an individual's unique decision-making context.

Each of us has our own distinct set of influences, circumstances, and core beliefs that drive our healthcare decision-making. One important variable is perceived self-efficacy (PSE), an element of Bandura's Social Cognitive Theory (SCT)1. PSE focuses on a patient's personal confidence beliefs about his or her capacity to undertake specific health behaviours that may lead to desired outcomes. In the context of chronic illness, PSE is predictive of quality of life and successful disease self-management2. Because self-efficacy beliefs are modifiable and can impact health status, motivation levels, goal achievement, adherence and persistence, intervention approaches that focus on boosting PSE can empower patients and improve chronic disease outcomes. Indeed, self-management programs designed specifically to improve PSE by incorporating self-confidence, goal-setting or control-enhancing strategies can produce more favorable outcomes for patients than with standard interventions3.

This talk will highlight examples where PSE and human connections are used in practice across a variety of contexts including maternal health, type 2 diabetes, and chronic kidney disease.

Automated decision making continues to be used for a variety of purposes within a multitude of sectors. Ultimately, what makes a ‘good’ explanation is a focus not only for the designers and developers of AI systems, but for many disciplines, including law, philosophy, psychology, history, sociology and human-computer interaction. Given that the generation of compliant, valid and effective explanations for AI requires a high-level of critical, interdisciplinary thinking and collaboration, this area is therefore of particular interest for Web Science. The workshop ‘Explanations for AI: Computable or Not?’ (exAI’20) aims to bring together researchers, practitioners and representatives of those subjected to socially-sensitive decision-making to exchange ideas, methods and challenges as part of an interdisciplinary discussion on explanations for AI. It is hoped that this workshop will build a cross-sectoral, multi-disciplinary and international network of people focusing on explanations for AI, and an agenda to drive this work forward. exAI’20 will hold two position paper sessions, where the panel members and workshop attendees will debate the following key issues in an interactive dialogue: The sessions are hoped to stimulate a lively debate on whether explanations for AI are computable or not by providing time for an interactive discussion after each paper. The discussion will uncover key arguments for and against the computability of explanations for AI related to socially-sensitive decision-making. An introductory keynote from the team behind the project PLEAD (Provenance-Driven & Legally Grounded Explanations for Automated Decisions) will present use cases, scenarios and the practical experience of explanations for AI. The keynote will serve as a starting point for the discussions during the paper sessions about the rationale, technologies and/or organisations measures used; and, accounts from different perspectives – e.g. software designers, implementers and those subject to automated decision-making. By the end of this workshop, attendees will have gained a good insight into the critiques and the advantages of explanations for AI, including the extent in which explanations can or should be made computable. They will have the opportunity to participate and inform the discussions on complex topics about AI explainability, such as the legal requirements for explanations, the extent in which data ethics may drive explanations for AI, reflections on the similarities and differences of explanations for AI decisions and manual decisions, as well as what makes a ‘good’ explanation and the etymology of explanations for socially-sensitive decisions. exAI’20 is supported by the Engineering and Physical Sciences Research Council [grant number EP/S027238/1]. We would like to thank the organizers of the Web Science 2019 conference for agreeing to host our workshop and for their support.

The purpose of STAIDCC20 workshop is to bring together a mixture of inter-disciplinary researchers and practitioners working in defence, cybercrime and cybersecurity application areas to discuss and explore the challenges and future research directions around socio-technical AI systems. The workshop will showcase where the state of the art is in socio-technical AI, charting a path around issues including transparency, trustworthiness, explaining bias and error, incorporating human judgment and ethical frameworks for deployment of socio-technical AI in the future.

This keynote will explore the broader issue of using socio-technical artificial intelligence (AI) systems in criminology for responding to cybercrime and cybersecurity issues. It will focus upon the importance of matching the delivery of AI with the scientific (technical) claims for it within a socio-political world. By drawing upon research into cybercrime and cybersecurity (including recent ransomware research), the talk will discuss the realities, the strengths and weaknesses, of using AI with regard to attribution and investigating cybercrime, and also preventing attacks to systems. It will argue that the meanings, logic and understandings of AI systems differ across disciplines which can result in significant differences in expectations. The broad conclusion is that because of this an interdisciplinary approach needs to be taken and that AI it is not a silver bullet. AI systems may be useful, for example, in responding to some cybercrimes, but not others, or effective in addressing stages of a cybercrime event, such as preventing malware infection; and even then, only with some major caveats. More importantly, is the recognition that AI cannot actually make hard decisions, but it can reasonably inform aspects of the decision-making processes of practitioners, professionals, policy makers and politicians who are mandated to make them. It is not only important to match the delivery of scientific claims with consumer expectations in order to maintain public confidence in the public security sector, but also because an arms races is developing as offenders are also beginning to employ AI in a number of different ways to help them victimize individuals, organisations and nation states [1].

The first part of this talk will draw upon existing examples to explore the general issue of using socio-technical AI systems to deal with crime and policing in a risk society [2] [3], before identifying some of the additional challenges presented by AI and cybercrime and cybersecurity [4] [5]. The second part will look at the methodological and socio-political problems of delivering science solutions within a socio-political world. The third part will conclude by discussing the practical realities, strengths and weaknesses, of using AI regarding attribution and investigating cybercrime, and preventing attacks to systems.

In today's online forums and marketplaces cybercrime activity can often be found lurking in plain sight behind legitimate posts. Most popular criminology techniques are either manually intensive, and so do not scale well, or focus on statistical summaries across websites and can miss infrequent behaviour patterns. We present an inter-disciplinary (computer science, criminology and conservation science) socio-technical artificial intelligence (AI) approach to information extraction from the long tail of online forums around internet-facilitated illegal trades of endangered species. Our methodology is highly iterative, taking entities of interest (e.g. endangered plant species, suspects, locations) identified by a criminologist and using them to direct computer science tools including crawling, searching and information extraction over many steps until an acceptable resulting intelligence package is achieved. We evaluate our approach using two case study experiments, each based on a one-week duration criminology investigation (aided by conservation science experts) and evaluate both named entity (NE) directed graph visualization and Latent Dirichlet Allocation (LDA) topic modelling. NE directed graph visualization consistently outperforms topic modelling for discovering connected entities in the long tail of online forums and marketplaces.

We are moving towards a world in which data have a life beyond the individual; where the value and potential of data are ever-changing as technological developments bring new possibilities. This immortality of data raises new ethical and societal issues that have not yet been fully articulated, and consequently we are unprepared to deal with. The challenges of dealing with large volumes of personal data are increasingly apparent in many fields of practice, although they may manifest in different ways. This workshop brings together participants from diverse domains to provide an opportunity to articulate across disciplines and stakeholders the commonalities in the issues that arise. We aim to stimulate debates about how we might work collaboratively to anticipate, manage and prevent future issues. The presentations and discussions will be used to prepare a summary white paper/symposium briefing document to be disseminated more widely.

The Web has been the subject of compelling biological metaphors that liken it to an evolving ecosystem. Analogies of this kind could benefit from further theoretical and empirical examination. Evolutionary and cognitive approaches provide not only a powerful theoretical framework to address this theme, but also a heritage of robust analytic tools that can help to quantify complex and subjective social and technological phenomena. The goal of this workshop will be to discuss how evolutionary approaches can inform our understanding of the Web at present, as well as methodological challenges and opportunities to shape its evolution into the future.