The DSI uses its strategic resources to support projects of DSI network members that meet the objectives of DSI.
Beyond Anthropomorphism? The AI - Machine - Human - Animal continuum
Applicant: Suzanne Tolmeijer
From religious idols to important buildings, humans have had relations with non-human entities for millennia. The most important of these relationships often result in radical anthropomorphisation of these entities, lifting the relationships towards being on par with relationships with other humans. Pets, vehicles, and now also computers can, under specific circumstances, be the object of deep feelings of humans. In the case of computers, this is known as the ‘computers as social actors’ paradigm.
How does anthropomorphic and zoomorphic artificial intelligence (AI) fit into this continuum of human-nonhuman relationships? In human-computer interaction, using anthropomorphism is characterized by some as seamless, a characteristic seen as desirable. Others instead propose “perfect seams”, an approach where the non-human counterpart explicitly retains their own machine strengths and exposes their own non-human weaknesses.
This project proposes to reverse the lens of inquiry into anthropomorphic AI. Instead of looking at whether AI that has human characteristics is perceived as more trustworthy, it asks how the anthropomorphizing of AI changes pre-existing relationships that humans have with machines, such as cars and planes, that are fitted with AI components.
The project will focus on the temporal evolution of the bond between human operators and the AI assistant embedded in planes, and explore facets such as alignment of the temperament of the machine, signalised by voice, and manifested in physical behaviour.
The DSI supports this project together with the Centre for AI and Digital Ethics (University of Melbourne).
Journal of Quantitative Description: Digital Media
Applicant: Dr. Andy Guess, Dr. Eszter Hargittai, Dr. Kevin Munger
The Journal of Quantitative Description: Digital Media is an open access, peer-reviewed scholarly journal launching in 2021 hosted on the University of Zurich's HOPE platform supported in its launch by the Digital Society Initiative. All articles are freely available online immediately upon publication and we plan no submission or publication fees for at least the first two years, but hopefully longer.
The journal publishes quantitative descriptive social science. It does not publish research that makes causal claims.
The production of descriptive knowledge is currently undersupplied in quantitative social science. As John Gerring documents in the case of political science in his 2012 article "Mere Description", scholarship in the postwar era has seen a steady trend from description to causality. We applaud developments in method that discredit spurious causal research; this has led to a much-needed advance in rigor for claims of causation. As this shift illustrates, social science has a momentum that takes years to redirect, and graduate training, journal space, prestige, and grant funding have all been shifting away from quantitative description. Our hope is that this journal is the beginning of another course correction.
Descriptive knowledge is necessary for the following steps in the social science process:
Hypothesis generation: Trivially, we need to know what is before we can derive hypotheses about why it is or what it does. Too often, experiments are designed without first establishing the prevalence of the causes or effects being studied.
Topic importance: We want to study the most important questions, but relying on the intuitions of social scientists about "importance" is baldly unscientific. Quantitative description offers a framework for rigor.
Generalizability: The goal of many social scientists is to create "generalizable" knowledge. There are open questions about how best to do this, but an essential component of any generalizability project is the knowledge of how the target context differs from the known contexts. This knowledge is quantitative description, which thus serves as a complement to causal knowledge, enabling its applicability to novel contexts.
The journal has a stellar international Advisory Board and Editorial Board representing several disciplines supporting its mission. The founding co-editors are Andy Guess, Political Science, Princeton University, former DSI Fellow; Eszter Hargittai, Communication and Media Research, University of Zurich, organizer of the DSI Book Club; and Kevin Munger, Political Science, Pennsylvania State University.
Workshops of the Computational Methods Working Group
In order to train scientists in Computational Methods and to discuss problems in the application of these methods in an interdisciplinary way, young researchers from the Institute for Communication Science and Media Research (IKMZ), the Research Institute Public and Society (fög) and the Institute for Political Science (IPZ) have founded a Computational Methods Working Group. The group consists of doctoral and post-doctoral students and is aimed explicitly at young researchers. The Working Group has two goals: On the one hand, it wants to offer young researchers methodological training for the application of digital methods. On the other hand, the Working Group aims at networking scientists who work with computational methods in an interdisciplinary way. For further education and networking the working group organizes two events in 2020: a workshop on "Text Classification in Python" (October 2020) and a conference on "Young Scholars in Computational Social Science" (November 2020). For more information, please visit the website of the Working Group.
The DSI supports this project together with the Graduate School of the University of Zurich.
Digital History Lab
Applicants: Prof. Dr. Monika Dommann, Prof. Dr. Martin Dusinberre, Prof. Dr. Gesine Krüger
The Department of History’s Digital History Lab (DHL) will tackle the crucial issue of how contemporary society should understand the historical emergence and evolution of digital societies. It will analyse a history of digital data, including its cultural and epistemological dimensions, while focusing also on societal issues (surveillance, privacy, cultural properties, public access, indigenous knowledge, citizen science, and the public understanding of academia). Drawing on existing digital history expertise in the Department of History, especially in Media History, History of Knowledge, E-learning, and digital dissemination, it will be a forum for sharing best practices, so as to offer practical and intellectual support to future projects. In these ways, the DHL will be a focus for research, teaching and engagement activities, offering scholars a space in which to reflect on and practice history in the digital age.
The DSI supports the Digital History Lab together with the Department of History.
Toolbox for digital work with manuscripts
Applicants: Prof. Dr. Marianne Hundt und PD Dr. Gerold Schneider
As part of a credit project, a toolbox for digital work with manuscripts will be developed, which will be tested and evaluated in a subsequent project seminar on diary manuscripts from the 18th century. The central element of the toolbox is an annotation tool that provides TEI and XML components for the digital edition of manuscripts with a user-friendly interface. The editor should be flexible enough to be used in other courses and research projects.
Project-Website: Toolbox for the Digital Annotation and Analysis of Manuscripts
The DSI supports the development of the editor during 6 months.
Applicant: Prof. Dr. Milo Puhan
The “SUN-sitive” project, based at the UZH Epidemiology, Biostatistics and Prevention Institute aims to develop a web-based approach to sensitise healthcare consumers about skin cancer prevention, mitigate tobacco-related skin damage and promote overall skin health, integrated in a primary care context.
The SUN-sitive web-app will be built on the infrastructure of a generic and highly flexible online research platform, developed by our institute. App access will be available to eligible participants on tablets in the waiting rooms of the UZH travel clinic. Once accessed, users will be asked to enter a few health-related data, followed by an automatically and instantly generated digital preventive consultation, tailored to the information that has been provided by each user. Based on strong theoretical foundations, its content will entail behavior change techniques and interactivity, targeting user attitudes, normative beliefs, self-efficacy, knowledge and behavioral intentions regarding exposure to UV radiation (sun, sunbeds) and tobacco use.
Our project aims to evaluate the potential impact of that approach on key prevention-relevant outcomes (e.g. intentions to protect from UV-radiation; awareness on the consequences of tobacco on skin appearance), as well as assess the feasibility and acceptability of digital prevention in a primary care setting. We aim to apply the knowledge gained from the development and evaluation of SUN-sitive to further adjust and improve its underlying platform, which is designed to be applicable across health domains and research projects. Our broader vision targets the sustainability of eHealth and digital prevention within primary care.
The DSI supports this project together with the EBPI and the Béatrice Ederer-Weber Foundation.
Die Virtuelle Universität in Zeiten von Corona
Workshop of the Digital Society Initiative (DSI) in cooperation with the Center for Higher Education and Science Studies (CHESS) of the University of Zurich (17.-18.09.2021)
„Endlich wieder einmal eine Veranstaltung im Präsenzmodus!“ – das dachten wohl alle der rund 50 nach strengen Corona-Sicherheitsvorkehrungen handverlesenen Teilnehmer*innen des zweitägigen Workshops. Das mehrfach verschobene Festkolloquium für den Philosophen Walther Ch. Zimmerli, DSI-Fellow und CHESS-Mitglied, sollte sich im Nachgang zu seinem 75. Geburtstag mit der „rekursiven Digitalisierung“ der Hochschulen befassen und konnte nun, wie erhofft, live stattfinden. Dass Academia nach wirklichem Gedankenaustausch mit realen Personen in Echtzeit geradezu dürstet, liess sich nicht zuletzt an der Intensität der Diskussionen über die von hochkarätigen Referent*innen aus dem In- und Ausland vertretenen Thesen ablesen.
Also alles beim Alten und Rückkehr zum analogen universitären „business as usual“? – Weit gefehlt! Die vorherrschende, von Zimmerli in Auseinandersetzung mit Wilhelm von Humboldt prägnant formulierte Auffassung war, dass die Pandemie eine „disruptive Innovation“ ausgelöst und Spuren hinterlassen habe, die die Hochschulen auch in Zukunft prägen werden. Fortschritte in der Digitalisierung von Lehre, Forschung und Entwicklung, die sonst wohl Jahrzehnte gedauert hätten, wurden, der Not gehorchend, nun innert Monatsfrist erzielt (Gabriele Siegert, Vizerektorin Lehre an der UZH, und Michael Hengartner, Präsident des ETH-Rates). Dass allerdings- vielleicht im Gegensatz zu den Fachhochschulen (Andreas Müller, Hochschule Kempten) - eine vollständige Digitalisierung gerade in der Universität weder möglich noch wünschenswert sei, wurde angesichts der Unabdingbarkeit des „menschlichen Antlitzes“ in der Gemeinschaft der Lehrenden und Lernenden (Jan-Hendrik Olbertz, ehemaliger Präsident der Humboldt- Universität zu Berlin) schnell klar, selbst wenn die erdrückende Vielfalt neuer Herausforderungen und Theorieansätze (David Bosshart, langjähriger Geschäftsführer des GDI) das auch jenseits der falschen Alternative von analog und digital (Markus Christen, Geschäftsführer der DSI) zuweilen nahezulegen scheint.
Die zwar zurückhaltend formulierte, aber inhaltlich brisante Frage, ob heute nicht schon Datenbanken wie etwa Springerprofessional so etwas wie virtuelle Universitäten darstellten, wurde eher skeptisch beantwortet (Dieter Beste, Düsseldorfer Journalist), während der Vorschlag, in der Ingenieurausbildung gerade in Zeiten der Digitalisierung auf alte Muster wie etwa die zünftische „Walz“ der Handwerker zurückzugreifen (Engelbert Westkämper, ehem. Direktor des Fraunhofer-IPA Stuttgart), auf starkes Interesse stiess. Dass bei aller digitalen Virtualisierung in Universitäten ebenso wie in grossen Unternehmen nun erst recht das Prinzip „Geist braucht Raum“ gilt, lässt sich an moderner Hochschul- und Unternehmensarchitektur zeigen (Gunter Henn, Architekt München und Berlin). Und wirklich beruhigend war zu sehen, wie kreativ sich Nachwuchsforscher*innen in ihren PhD-Projekten bei allen Einschränkungen auf die Coronasituation eingestellt haben (Johanna Jud, Alexandra Zoller und Kenan Hochuli, Universität Zürich).
Was bleibt, und wie geht es weiter? Im Abschlussgespräch zu den Zukunftsperspektiven waren sich Gerd Folkers, ehemaliger Präsident des Schweizer Wissenschaftsrats, und Walther Ch. Zimmerli, einig, dass die Digitalisierung in besonderem Masse die Kompetenz der reflektierenden Urteilskraft erforderlich mache und dass dazu gerade die Universitäten institutionelle Freiräume etwa in Gestalt von Institutes of Advanced Study schaffen und pflegen müssten. Es überrascht nicht, dass Sebastian Bonhoeffer, neuer Direktor des Collegium Helveticum, und Bernd Kortmann, Direktor des Freiburger Institutes of Avanced Study FRIAS, diese Meinung teilten.
What is the proportion of robots in Tweets? – BOTS
Applicant: Dr. Oliver Grübner
Read the project report here (PDF, 680 KB).
Nearly half of the world’s population now use social media for some form of communication. However, it is unclear how many bots, that is, accounts that are operated by non-human robots, influence the public conversation in social media in general and discourse on public health topics in particular.
The project seeks to advance existing bot detection methodologies in social media by applying a combination of geographic trajectory and emotion analysis to explore bot detection from Twitter that later could help in improved public health surveillance.
BOTS will deliver crucial knowledge about the nature of bot tweets (type, geographic trace, and emotive composition) that can be investigated in relation to bot tweets’ volume by topic, compared to user-generated content and their impact on users’ subsequent activity. Furthermore, having a better understanding about bot activity in social media (qualitatively and quantitatively) will help to more precisely (with and without bot accounts) identify geographic areas (and populations) exhibiting high concentrations (hotspots) of specific emotions or stress, as well as track the spread and distributional dynamics of stress related content (e.g. social contagion) in real time.
Team members include Oliver Gruebner (PI, Department of Geography), Dr. Markus Wolf (Department of Psychology), Dr. Fabio Rinaldi (Institute of Computational Linguistics). Significant contributions will be made by Dr. Suzanne Elayan and Dr. Martin Sykora (both Loughborough University, UK), Dr. Yulin Hswen (Harvard Medical School), and Dr. Nason Maani (Boston University School of Public Health).
The DSI supports this project.
An empirical analysis of access to federal administration data in accordance with the principle of public access
Applicant: Prof. Dr. iur et Dr. med. Kerstin Noëlle Vokinger
Translated form German.
Aim of the study: To empirically examine all Federal Administrative and Federal Court decisions (2006 - 2019) that have had access to official documents within the meaning of the Public Disclosure Act on the subject of dispute with regard to the following questions: How has the number of "BGÖ complaints" developed between 2007 - 2019? Who was a party in such court proceedings on the side of the authorities? Who was a party in such court proceedings on the side of the society (journalists, private persons, lawyers, companies, interest representatives, scientists)? What was the specific subject matter of the dispute (in particular fees, applicability of the Law on Publicity, qualification as an official document)? What was the outcome of such court decisions?
In addition, we investigated the "citation behavior" of the Federal Administrative Court and the Federal Supreme Court. Case-law and doctrine interact with each other - in making their decisions, courts rely, among other things, on doctrinal findings (and vice versa). In the case at hand, we were specifically interested in the relationship between the Federal Administrative Court and the Federal Supreme Court in which materials, teaching, case law and laws are cited and how this has developed since 2006. In this context, we also examined whether differences between the two courts can be identified.
- The number of "BGÖ-complaints" before the Federal Administrative Court and the Federal Supreme Court has tended to increase since the law on publicity came into force.
- On the administrative side, SECO, Swissmedic and the FTA were most frequently involved as parties.
- On the society side, journalists were most frequently involved, followed by interest groups (clubs, associations and unions), private individuals and companies. The courts never anonymized the names of interest representatives and anonymized the names of journalists and entrepreneurs in 40% of cases and private individuals in 60% of cases.
- Most frequently, both courts had to decide whether in a specific case the public interest in access to official documents outweighed the private sphere of third parties. Other frequent disputes were whether in the specific case the document qualified as an official document and whether the BGÖ was applicable. If personal data was requested, it was primarily data of public figures. Access to factual data was also frequently requested.
- 73.4 % and 55 % of the complaints before the Federal Administrative Court and the Federal Supreme Court were (partially) approved. This is a remarkably high rate.
- Most frequently, both courts cited legal provisions in their deliberations, followed by earlier rulings, doctrinal works and materials. The number of doctrinal works cited has decreased in recent years, while the citation of earlier judgments has increased. One possible reason is that, especially in recent years, the scientific publications on BGÖ are manageable. It is noteworthy that the courts have not drawn on general literature on constitutional and administrative law.
- The Federal Administrative Court and the Federal Supreme Court most frequently cited Art. 7 BGÖ (exceptions to the principle of publicity) followed by Art. 5 BGÖ (definition of which documents qualify as official documents).
In our opinion, the results are the first indications that such an empirical analysis of court decisions could be suitable as an evaluation element for a review of the effectiveness of laws within the meaning of Art. 170 BV - understood as the analysis of whether the implementation of a norm or decree has proven to be successful and whether the objectives have been achieved. The present analysis shows that with such an empirical approach it is possible to determine, for example, the increasing number of BGÖ complaints or the concrete legal provisions that are frequently cited as well as the concrete circumstances for their citation.
The empirically based study results show that the law on publicity is not only theoretically an important legal basis for the implementation of the constitutional principle of transparency, but that there is also a real need in society to obtain information on state activity.
Here (PDF, 419 KB) you can read the paper.
DSI supports this project together with the RWI.
Digital visual information therapy (DigiVisIn) for orofacial pain
Applicants: PD Dr. Dominik Ettlin, Prof. Gerhard Schwabe, Dr. Mateusz Dolata, Dr. Markus Wolf
Communication of symptom-related information in nonprofessional terms plays an important role in disease management. The aim of information therapy is to inform patients about aspects of health maintenance and illness development so that they can make informed decisions based on a profound understanding. Imaging in medicine, for example, is not only used for diagnostics (e.g. intrauterine development of the fetus), but also to promote human understanding of complex processes. The didactic value of images is illustrated by the proverb: "A picture tells more than a thousand words".
Persistent complaints in the orofacial area (mouth and face) are mostly due to molecular changes in muscles and nerves, which cannot be visualized by imaging methods. In addition, psychosocial stressors commonly influence somatic symptom perception and vice versa, bodily symptoms negatively impact on psychological well-being. In current clinical practice, verbal explanations and hand-drawn sketches provide information regarding biological mechanisms and bio-psychosocial interactions.
The objectives of this project are the development and implementation of an interactive, digital tool to support patient-centered information therapy (DigiVisIn). DigiVisIn visualizes expert knowledge by means of biological illustrations and graphical models, which are adaptable to the individual needs for information therapy. The structure is modular so that different aspects are presentable based on requirements. Due to the didactic focus on comprehension by layperson, also specialists from various medical and paramedical disciplines with little expertise in the area of orofacial problems are able to convey the necessary information at a high quality level. In particular, the projects aim at empowering students in conveying complex biopsychosocial relationships in a way that they are easily understandable for laypersons. Once approved by the ethics committee, the application will be tested in a clinical context. The DSI application serves as financial bridge for the preparation of an Innosuisse application.
The DSI application serves as financial bridge for the preparation of an Innosuisse application.
Innovating digital analysis methods for patient-centered health research – development of a text processing pipeline in the Swiss Multiple Sclerosis Registry
Applicant: PD Dr. Viktor von Wyl
Patient-centered research demands that patients’ views are afforded greater attention. In particular, research questions, but also study outcomes, should be better geared towards addressing patients’ health needs. Along the same lines, study participants are increasingly less willing to just “tick boxes” in research studies, but rather would like to “tell their story”. However, standard methods of data collection and analysis (e.g. those relying on highly structured questions in health surveys) often fail to adequately capture a person’s individual situation in sufficient detail.
The goal of this project, which is funded by the Digital Society Initiative and the Swiss Multiple Sclerosis Society, is the development of digital tools to strengthen patient-centeredness in epidemiological, long-term study. We aim to develop an automated analysis pipeline for unstructured text information, with the ultimate goal to detect and quantify novel health topics. For this purpose, 3,000 anonymised free text entries from the Swiss Multiple Sclerosis Register (SMSR) were analysed. From this text data, key words and phrases were automatically extracted which are characteristic for different life topics (e.g. work, well-being, as well as MS-specific topics). The underlying dictionary was developed together with MS experts and people with MS. The correct keywording of the free text entries with the dictionary is now being tested in a pilot trial on a random basis together with 20 people with MS. For this purpose, an app has been developed with which volunteers can validate the text entries online. A publication on these analyses and the planned validation is in preparation.
The tools developed in this project will now be used to analyse a recently completed text-based survey of 750 people with MS on the most important events in life with MS. On the one hand, these entries will be scientifically processed to characterise typical MS courses. On the other hand, the entries will be catalogued and made available to people with MS in an anonymous, searchable form, because these text entries contain valuable experiences.
This project is supported by the Digital Society Initiative and the Participative Academy of Science of the University of Zurich, as well as Switzerland. MS Society.
The Integrative Data Management Portal for Longitudinal Human Development Data: An innovative contribution to research infrastructure at the University of Zurich
Applicants: Dr. Franz Liem, Prof. Dr. Mike Martin, Dr. Susan Mérillat, Dr. Christina Röcke
The way in which scientific research in the social sciences is conducted has fundamentally changed with the growing influence of digitalization and technology and results in increasing demands on data management in order to facilitate the sustainable use of data throughout their lifecycle. In the context of studying human development, the sustainable use of data is of particular relevance because the acquisition of longitudinal data, which are needed to draw conclusion about developmental trajectories, is usually very time- and cost-intensive, rendering the resulting datasets enormously valuable.
With the support of the Digital Society Initiative we developed a comprehensive concept for an up-todate data management portal, which – as an innovative infrastructure – embeds a modern data storage infrastructure in a bigger framework of services (e.g. education, consulting, etc.), technologies and outreach functions and, therefore, strives for a more efficient use of the data stored in the portal (project goal 1). Implemented as an interdisciplinary network of specialists, this portal is intended to offer specific, high-quality support to researchers and free up resources for the actual research. A prototype of the technical implementation for data storage and data processing has been implemented (project goal 2) and is currently being tested.
Based on the concept and the technical prototype, we will survey requirements and needs of future portal users by the end of 2019 in order to optimize functionality and user friendliness of the portal at an early stage and to be able to realistically prepare the cash budget (project goal 3). The application for acknowledgement of the data management portal as technology platform is planned for 2020.
The DSI supports this project together with other sponsors.
Workshop on Text Analytics for Health: Applications and Implications
On September 12-13, 2019 the DSI challenge area Health organized an international workshop on “Text analytics for health: Applications and Implications”. During these intensive 2 days, 7 international experts on computational linguistics for health (Lawrence Hunter from Denver Colorado, Goran Nenadic from Manchester UK, Pierre Zweigenbaum from Paris France, Ryan Boyd from Lancaster UK, Michael Krauthammer from Unispital Zurich, Martin Volk and Fabio Rinaldi from the Institute of computational linguistics UZH) shared their know-how and experiences with 30 participants. They listened and gave valuable inputs to all participants who presented their own projects. The work in subgroups allowed the identification of future opportunities and challenges for text analytics in the health domain at the UZH, and to envision future collaborative projects.
Mike Martin, a DSI director supporting the challenge area Health, concluded the meeting by bringing up different possibilities to leverage the UZH text analytics expertise. Several disciplines are working on free texts in health and a lot of information could be extracted at a multiscale level. Assuming more resources would be granted to text analytics, priorities would have to be set regarding which topics, methods and structures to pursue. Larry Hunter confirmed the need to cope with a continuous change and improvement in technologies, the difficulty to find experts, the necessity to build skills, organize training programs, and optimize IT architecture.
To sum up, we learnt and exchanged extensively in this workshop, and built the foundations for more development, collaboration and projects for text analytics in the health domain at UZH.
Medical and Regulatory Characteristics of Approved Artificial Intelligence/Machine Learning-Medical Devices in Switzerland, the EU, and US and Health Policy Implications
Applicants: Prof. Dr. iur. et Dr. med. Kerstin Noëlle Vokinger, LL.M., Thomas J. Hwang, AB
The US Food and Drug Administration pointed out in spring 2019 that new regulatory mechanisms are necessary for AI/ML products. We analyzed the current regulatory requirements for AI/ML medical devices in the U.S. and then specifically investigated which AI/ML medical devices (indication, approval year, approval route) have been approved in the U.S. to date. Based on these analyses, we have identified implications for what regulatory measures would be appropriate to simultaneously strengthen the innovation of AI/ML and ensure safety for patients. Among other things, we have come to the conclusion that it is important to take into account in the regulation that - in contrast to traditional software - certain AI/ML medical devices have the ability to develop further after approval (so-called "adaptive software"), whereby the outcome of this further development is not always clear at the time of approval, which is why we also speak of the necessity of "lifecycle regulation". In our study we propose concrete regulatory measures.
Our study with the final title "Lifecycle Regulation of Artificial Intelligence- and Machine Learning-Based Software Devices in Medicine" was accepted by the international peer-reviewed journal "JAMA" (Journal of the American Medical Association), one of the world's most respected medical journals. The online publication is scheduled for November 21, 2019, the print publication for December 17, 2019.
The DSI supported this project.
Visual analysis of audiovisual data
Applicants: Prof. Dr. Barbara Flückiger, Prof. Dr. Renato Pajarola
The visual video annotation software VIAN, developed within the scope of the FilmColors.Bridging the Gap Between Technology and Aesthetics project, funded by the European Research Council (ERC) with an Advanced Grant, is a comprehensive platform for the analysis and visualization of film and video. VIAN not only enables the annotation of sequences, it can also perform complex evaluations. With Deep Learning Tools it is possible to separate the foreground and background for the analysis as well as to recognize figures and gender. Thanks to the detailed recording, VIAN can arrange the sequences accordingly and visualize them dynamically and interactively in various graphs.
With the support of the Digital Society Initiative, Digitale Lehre und Forschung and the Citizen Science Competence Center, VIAN has been developed into a comprehensive crowd sourcing client with an offline tool for video analysis and a web portal for publishing and preparing the results. A usability study with a selected group of external users is currently underway and the general release is planned for the coming weeks.
VIAN was developed by Gaudenz Halter together with the Visualization and MultiMedia Lab (VMML) led by Prof. Dr. Renato Pajarola at the Institute of Computer Science as part of the ERC-Advanced Grant project "FilmColors". With the support of the DSI, Digitale Lehre und Forschung and the UZH Competence Center Citizen Science, VIAN was brought into a platform that can be used by a broad audience.
International Cyber Security Summer School (ICSSS 2019)
Applicant: MLaw Salome Stevens
With the support of the DSI, Ms Stevens participated in this years International Cyber Security Summer School held in The Hague. The Summer School brought together researchers and experts from all disciplines with the aim of fostering an interdisciplinary understanding of Cyber Security. Quickly it became clear that an effective Cyber Security Strategy demands the interdisciplinary cooperation of technical, legal and policy people, but also of psychologists and sociologists in all processes from design to reactive security. Interdisciplinary communication can however be challenging as experts are often reluctant to think across the expertise, they feel comfortable in. The Summer School has shown how important it is to fight these interdisciplinary boundaries. This can be achieved through projects that aim to facilitate interdisciplinary communication and overcome such inhibitions. One approach could be interdisciplinary training. No one likes to speak about things she or he does not know anything about. It is therefore important that experts are provided with a basic knowledge of the core interests of other disciplines in order to strengthen confidence in an interdisciplinary exchange. Experts that already have an interdisciplinary knowledge of Cyber Security can also be put in place as “translators” in order to facilitate communication and avoid misunderstandings. It is only together that we can work on a balanced and goal oriented Cyber Security Strategy that guarantees for more security for everyone in Cyberspace.
More information on the International Cyber Security Summer School: https://www.summerschoolcybersecurity.org/
The DSI supported the participation in the summer school.
Applicants: Prof. Dr. Leonhard Held, Dr. Eva Furrer
The objective of the UZH Reproducibility Day was to provide researchers information on issues of reproducibility, on opportunities for solutions and on offers at UZH. They were invited to take the plunge and practice with experts from CRS in hands-on workshops. The day consisted of presentations and workshops by experts from across UZH as well as invited speakers: Martina Grunow, managing editor of The International Journal for Re- RW Views in Empirical Economics and Nathalie Le Bot, senior editor at Nature gave the publisher's perspective on replication and reproducibility.
All presentations and downloads of the Reproducibility Day 2019 are on OSF.
The detailed program was:
This workshop was supported by the DSI.
Credibility of Conspiracy Theories
Applicant: Dr. Juliane A. Lischka
Conspiracy theories are a frequent topic in online discussions, have a long online distribution life and are almost globally available on digital platforms. The belief in conspiracy theories is often related to a decrease in trust in media and politics. The evaluation of conspiracy theories ranges from being thought provoking to insane. This project aims at understanding (a) how online users conceptualize conspiracy theories, (b) which cues are decisive for assessing conspiracy theories as credible and (c) in what way the belief in conspiracy theories determine future behaviour. Data is collected with online surveys and analysed with standard social science as well as machine-learning approaches.
The DSI supported this project with the funding of data collection costs.
APARIUZ Series Volume XX - "Digitization and Law"
Applicants: Dr. iur. Alexandra Dal Molin, Dr. iur. Anne Schneuwly, MLaw Jasna Stojanovic
Since the founding of the APARIUZ series in 1999 (Analyses and Perspectives of Assistants at the Law Institute of the University of Zurich), young legal scholars at the University of Zurich have published an annual anthology of articles on fundamental legal issues. This tradition has been maintained this year thanks to the generous support of the DSI and other sponsors: In June 2019, the 20th (anniversary) volume of the APARIUZ series entitled "Digitalization - Society - Law" was published by DIKE Verlag.
The rapidly advancing digital transformation is shifting the focus of our everyday lives and social interactions more and more from an analogue to a digital world. Digitalization is not only affecting society, however, but is also presenting our legal system with a multitude of new challenges. Against this background, a total of 27 authors and one guest author have examined various aspects of the complex interplay between digitalization, society and law in thematically wide-ranging articles from all fields of law. In particular, the following research questions were raised and answered with independent, convincing proposals for solutions: How does digitalization change our law? Can our legal system keep pace with digitalization or are adaptations necessary? What responsibility does the state have in this respect? What influence does the law have on digital progress? Does digitalization make our legal system superfluous?
In answering the research questions posed, the authors have also worked on an interdisciplinary basis, thereby pushing forward innovative ideas at the interface of jurisprudence and technology. Shortly after the publication of the printed work, the one-year APARIUZ project was concluded with a book launch, which provided the opportunity to establish valuable contacts with professors and practitioners.
The DSI supported this project together with other sponsors.
Online publication of the Zurich Tangram Corpus
Applicants: Prof. Dr. Elisabeth Stark, PD Dr. Wolfgang Kesselheim, Dr. Tanja Samardzic
The Zurich Tangram Corpus includes more than 30 hours of audio and video recordings consisting of Face-to-Face-interaction (Swiss varieties of High German) as part of a psycholinguistic experiment (based on the classical experiment by Clark & Wilkes-Gibbs 1986). The corpus, that was compiled by Yshai Kalmanovitch as part of his dissertation project, allows the investigation of processes regarding verbal and nonverbal accommodation, both within individual conversations and over a longer period. This opens up a wide range of research possibilities, ranging from phonetic issues regarding sound accommodation up to accommodation phenomena in gestures, facial expressions and looks. The aim of the project is to publish the transcribed audio- and video-data in a format (web interface) which allows queries as well as liveediting. The result of this Lighthouse-project is not only a widely usable, publicly accessible audiovisual corpus, but also a best practice documentation in which concepts for the development of a central linguistic research infrastructure at the University of Zurich are specified.
The DSI supported this project together with the Zurich Center for Linguistics (ZüKL) and the URPP Language and Space.
Digital Democracy Center
Applicant: Prof. Dr. Fabrizio Gilardi
The DSI matching funds were used to set up the Digital Democracy Lab. Specifically, the funds were used to hire a postdoc as well as two research assistants who helped to prepare a research grant proposal that was submitted to the SNF Digital Lives program. The proposal was accepted and we were awarded a grant of 236'355 Fr. over a period of 15 months, starting December 2019. Moreover, the postdoc and research assistants worked on building prototypes for the data collection and analysis routines that the Digital Democracy Lab will continue to develop in 2020. Therefore, the DSI matching funds have enabled us to kick-start the Digital Democracy Lab prior to obtaining the SNF funding.
The DSI supported this project for one year with the funding of positions of research assistants.
Advancing the Assessment of Dynamic Psychological Processes in Daily Life
Applicant: Prof. Dr. Mathias Allemand and national/international collaborators
Thanks to the generous support of the UZH DSI, an interdisciplinary research meeting on the progress in the measurement of dynamic processes using digital technology was held at the University of Zurich on November 29 and 30, 2018. The goal of the two-day research meeting was to bring together scientists with different backgrounds and expertise to share innovative technology-based models and "tools" for capturing dynamic psychological processes and joint research projects. Different aspects of the measurement of dynamic processes were discussed in plenary sessions. Topics such as personality diagnostics, ubiquitous social networking, measurement burst designs, ecological momentary assessment, crowd sensing by means of mobile sensors and applications and machine learning were discussed. Moreover, we considered the opportunities (e.g., product design of applications) and challenges (e.g., age-related challenges) of digitally supported assessment (more information on the program of the research meeting (PDF, 1 MB)). The importance of measuring dynamic processes using digital technologies is particularly important in application areas such as Digital and Mobile Health. With the support of the UZH DSI, a guest lecture on the topic was held on December 7, 2018 (more information can be found here (PDF, 65 KB)).
The DSI supported this project together with other partners.
Recent article on the topic: Mobile Data Collection: Smart, but Not (Yet) Smart Enough (PDF, 128 KB)
Workshop "The Humanitarian Cyborg"
Applicant: Prof. Dr. Nikola Biller-Andorno, Ning Wang
The Cyborg Days is a 3-day workshop funded by the Graduate Institute (GRC), and hosted by the Institute for Biomedical Ethics and History of Medicine (IBME), with the support of the Digital Society Initiative (DSI). It took place on 26-28 November 2018, at the University of Zurich (UZH).
Thanks to the generous support of the DSI, the workshop brought together 8 scholars of diverse scientific backgrounds, and 5 practitioners from international organisations with expertise in technology and innovation, to gather in Zurich for in-depth discussions about the present and future of human-machine interactions. The unique combination of philosophical conceptions, scientific theories, the state-of-the-art technology showcase, and the policy and regulatory implications of real-world challenges, has made the event a catalyst of ideas, inspiring both scientific minds in the ivory tower and operational hands on the ground.
The workshop achieved its intended objectives with success. One distinctive added value was the incorporation of a special session on “The Humanitarian Cyborg” - a particular application of human-machine interaction in today’s digital world. The idea behind was that new and emerging technologies pervade our societies in multi-faceted ways and forms, sometimes even in the least explicit way, such as in the humanitarian space where, for centuries, human has been perceived as the essence of the very spirit and practice of humanitarianism. To portray and understand the real-world scenario of how technologies have been rapidly deployed in the humanitarian sector, increasingly replacing humans with machines, the special session featured practitioners and policy-makers from 4 international organizations, 2 UN agencies and 2 NGOs, to each present how human-machine interactions challenge their day-to-day operations.
The special session generated great public interests in two respects: 1) unlike robots and algorisms saturated in popular media, humanitarian cyborgs are rarely mentioned or even noticed, thereby exposing the audience to a new topic; and 2) unlike conventional academic workshops where science is the focus of discussion, policy and regulatory implications of science were being shedded lights upon, hence directing the audience to a different dimension of the relations between science, technology and society. In addition, the special session, as a key integral component of a newly-launched research project on “Value Sensitive Humanitarian Innovation: Integrating Values in the Humanitarian Use of Drones”, also officially kicked off the research project which successfully achieved funding from the Swiss Network for International Studies (SNIS), thanks to the continued support of the DSI.
The DSI supports this project together with the Graduate Campus and the Institute of Biomedical Ethics and History of Medicine (IBME).
Rethink Privacy! Towards a Harms-Focused Governance of Digital Traces
Applicants: Dr. Alfred Früh, Prof. Dr. Michael Latzer, Prof. Dr. Florent Thouvenin
Privacy is a key factor for individual and social well-being. In the digital age, ubiquitous data processing practices by businesses and government agencies and the abundant digital traces we knowingly or unknowingly leave behind affect privacy in various ways with consequences for individuals and society. The attempts to meet these challenges are heterogeneous and not based on a common understanding. Policymakers, for example, have reacted by enacting stronger privacy and data protection laws. Particularly the European Union (EU) has taken a comprehensive approach by enacting the General Data Protection Regulation (GDPR; effective May 2018). But the conceptual basis of the GDPR and other data protection laws, such as the Swiss Data Protection Act (SDPA), date back to the 1970s and are unable to fully meet the challenges of the digital age. To ensure that the processing of digital traces ultimately benefits individuals and society, we launch a research project that rethinks privacy with a synergetic combination of four perspectives: Philosophy (University of Lucerne), communication studies (UZH, IKMZ), law (UZH, ITSL), and technology (Università della Svizzera italiana). In addition, it collaborates with behavioral economics (ETH Zurich) on selected challenges. Governing privacy in the digital age in a meaningful way requires contributions from all these disciplines.
The project is structured in three parts. In the first part – Deconstructing Privacy – the project team explores the definitions, ascriptions, perceptions, and concepts of privacy as well as existing mechanisms to protect it. Combining different disciplines and defining a common language will result in a common understanding of the rich research in each field and enable an interdisciplinary synthesis. Upon these findings, the second part – Reshaping Privacy – starts out on the presumption that the processing of digital traces can be both beneficial and harmful, and that current regulatory and technical attempts have their limitations to successfully fight the actual harms, thereby curtailing important benefits. We explore three (potential) harms that are particularly important: manipulation, discrimination, and chilling effects (i.e., deterring or inhibiting certain actions or interactions due to, e.g., state surveillance or user-tracking by businesses). By analyzing these harms with an interdisciplinary approach, we develop a better understanding of how they affect individuals, groups, and society at large. Further, we will be able to develop criteria and benchmarks to separate the acceptable from the unacceptable, e.g., to distinguish manipulation as a harm from acceptable forms of influencing. In the third part – Governing Privacy – all disciplines will jointly devise governance arrangements that minimize the harms caused by the processing of digital traces while allowing the benefits to come to fruition. For each of the three harms, we will be able to draft novel, adequate, and workable governance recommendations that could have the form of technical solutions, legislative proposals, “green papers”, or calls for industry self-regulation including guidance. Based on a comparison of the governance recommendations for the three harms, we ultimately aim to draw up the foundations for a new governance framework for privacy in the digital age.
The DSI supported the grant application for the Rethink Privacy project together with the Faculty of Arts.
Conference "Sports, Data, and Journalism"
Applicant: Anil Özdemir
On October 25, 2018, the official Swiss Digital Day, the Center for Research in Sports Administration (CRSA) invited academics, data journalists, and experts from media and industry to the conference on Sports, Data, and Journalism at the University of Zurich to discuss the use of data science in sports.
At the sold-out event, over 100 participants from around the world listened to more than 35 interesting talks, participated in data science workshops, and exchanged ideas on the use of sports data. The conference was organized in an academic and data-driven journalism track. To allow an exchange of ideas and a discussion on methods, panel sessions, poster sessions, and workshops were offered to all participants.
Social scientists analyze sports data to test hypotheses, answer research questions, and develop new theories. Journalists ask similar research questions and analyze the same data. They might, however, use different methods, have different target audiences and publishing procedures. The conference provided a platform to discuss those different methods and the research questions asked. In some ways, both social scientists and data journalists were tackling very similar problems (e.g., judge biases in ice skating).
Although sports data is readily accessible, a key issue in the day to day work with sports data was that data came in structures that required heavy cleaning work before they could be analyzed. This opened a discussion on a potential sports data platform that would ensure data was made available in formats that made them easier to handle and analyze.
The multidisciplinary conference was very well received and we were asked whether there will be future conferences on the same topic.
The DSI supported this conference together with other partners.
Conference of the "International Humanitarian Studies Association"
Applicant: Ning Wang
With the generous support of the DSI in May 2018, a new research project “Value-Sensitive Humanitarian Innovation - Integrating Ethical Values in the Humanitarian Use of Drones”, hosted by the Institute of Biomedical Ethics and History of Medicine (IBME), successfully achieved a research grant from the Swiss Network for International Studies (SNIS) in July 2018. In August 2018, with continued support of the DSI and the IBME, Ms. Ning Wang, the principle researcher of the SNIS project represented the research team in the World Conference on Humanitarian Studies, in the Hague, Netherlands.
This conference is organised by the International Humanitarian Studies Association (IHSA) every two years. It is a world-renowned conference in the humanitarian sector, providing a platform for the different disciplines that address humanitarian studies on humanitarian crises caused by natural disasters, conflicts or political instability, and offering a venue where scholarly communities can meet and have dialogues with policy actors and relief and aid agencies. One of the major thematic focuses of 2018 was “The Impact of Technology and Innovation in Coping with, Responding to, and Transforming Crises”, within which a panel on “The Ethics and Technology in Humanitarian Settings” was chaired by Prof. Dr. Matthew Hunt, the Co-PI of the SNIS project, at McGill University, Canada.
During the conference, the SNIS team successfully established a number of new and strategic partnerships, many of whom will soon join the SNIS project as extended project partners. These novel partnerships not only include scholars and researchers who stand at the forefront of humanitarian studies, but also extend to funding organizations and policy think-tanks, all of which will play critical roles at different stages of the SNIS project. In sum, the SNIS team’s participation in the IHSA conference has proven to be fruitful and effective.
The DSI supported the applicant together with the Institute of Biomedical Ethics and History of Medicine.
Workshop "Digital Health and Shared Decision-Making"
Applicant: Dr. Corine Mouton Dorey
The workshop was a one-day multidisciplinary activity looking for the impact of digitalization on shared decision-making in healthcare. As digital health is an important challenge in a digital society, we had a lot more applications than the 20 planned with the graduate campus grant. Thanks to the DSI support we could increase the number of participants up to 30. The interactive format of the workshop was very well appreciated. The key success factors (KSF) can be summarized in 3 main points: First, we had talented speakers who bring their expertise in: Shared decision-making and autonomy (Prof. Schumacher, Uni Fribourg). Social systems theory and digital media (Prof. Vogd, Witten, Germany). Digital health (PD Dr. Christen, UZH). They interacted very generously during the day and took part in the subgroup works. This gave the possibility for PhDs and post-docs to develop fruitful discussion on their research topic. The second KSF was the diversity of people: experts in IT from UZH or ETHZ, in biomedical ethics or narrative literature. This made all discussions very rich. Interestingly, IT people were looking for more ethical perspectives, and the other non-IT attendants for more knowledge on algorithms. The third KSF was the work in three subgroups around the question of: Design a digital tool to improve shared decision-making in healthcare settings. The strategic thinking is different coming from philosophy, medicine or IT. However, each team came up with similar approaches that showed the high expectation for a positive digital influence on patient-physician relationships. Following the event, one sub-group decided to continue to work together on the digital tool they started to design. Some of the attendants are now working on a common position paper on digital health and shared decision-making. A follow-up meeting is planned before Christmas.
The DSI supported this workshop together with the Graduate Campus.
Consortium for H2020 proposal CHAnGE (Crowd Health Alignment and Greater Engagement)
Applicant: Dr. med. vet. Simon Rüegg
From 7 to 10 August 2018, Simon Rüegg from the Section of Epidemiology at the Vetsuisse Faculty convened a workshop to discuss ways to address One Health from a perspective of participatory governance. One Health (OH) is conceptualised as integrating decisions on human, animal and ecosystem health across sectors and disciplines, using systems theory and participatory processes. During the workshop the participants had lectures and exercises on personal construct psychology (PCP), while exploring in parallel opportunities and challenges of building a digital tool to facilitate system thinking and citizen participation in One Health governance. A sequence of drawing workshops, world café and mind-mapping sessions were used to construe the project. The core idea was to develop a smart phone enabled tool to elicit value systems associated to the different aspects of OH relying on PCP. The consortium decided to aim for an ERC Synergy grant with a project consisting of five working packages, but also explore alternative funding opportunities for fragments of the project. The working packages were, a package for transdisciplinary process management, a package for qualitative data collection and gaining a deep understanding of the landscape of values in which OH is situated, the development of a prototype of the app to be tested in sentinel populations, the roll out of the final app with options for user engagement and pathways from values to action for health improvement, and finally a package commending and customising the app for stakeholders at various levels of OH governance between international organisations and local users. The project is entitled Crowd Health Alignment and Greater Engagement (CHAnGE) and is currently involving eleven scientists form public health, animal health, food systems, sustainability science, health diplomacy, psychology, computer science, economics. It was interesting to see that all have professional trajectories involving both, quantitative and qualitative approaches, and were exceptionally agile in interdisciplinary conversation. The project plan will now be developed in more detail and submitted for funding within the next year.
Further details: https://www.vetepi.uzh.ch/en/research/CHAnGE.html
Together with Vetsuisse, the DSI supports this project.
Workshop "Cybersecurity, Industry and Ethics"
Applicant: MLaw Salome Stevens
Thanks to co-financing from the DSI, the applicant was able to participate in the workshop "Cybersecurity, Industry and Ethics" in Helsinki. The participation served as research for a book contribution entitled "Ethical Cyber-Defence for Companies", which is edited by the DSI Fellows PD Dr. Markus Christen and Dr. Michele Loi, among others. The workshop was helpful for the applicant's book project in various ways. The speakers' presentations provided a good overview of the ethical problems that cybersecurity actors are currently confronted with. In particular, questions relating to Ethical Cyber-Defence for Companies were discussed, which provide concrete clues for the research on the book chapter. In addition, by participating in the workshop, the applicant was able to broaden her understanding of the technical aspects of cyber attacks. Last but not least, participation in the workshop made it possible to establish contacts with cybersecurity actors and experts from research to whom it can refer in the course of the research.
Together with the CANVAS project, the DSI supported the project.
Brainhack Zurich 2018
Applicant: Dr. Franziskus Liem
The aim of the Brainhack was to network the neuroscientific community and to provide an introduction to open science ideas and tools in order to make neuroscientific studies more reproducible in the future. Besides the organized program, which consisted of a mixture of tutorials on tools such as Git, GitHub, Docker, analysis of DWI data and lectures on topics such as transparent and reliable science, neural networks and mobile data acquisition, an important part of the workshop was organized by the participants themselves. In spontaneous groups, topics such as data organization or analysis of longitudinal data were discussed. This format allowed a lively exchange between participants with different backgrounds and expertise and contributed to an open and inspiring atmosphere. More...
Together with the UFSP "Dynamics of Healthy Ageing" the DSI supported this hackathon.
Workshop "Philosophical questions on AI governance and law"
Applicant: Prof. Dr. Christoph Graber
The aim of the workshop on 15/16 February 2018 was to explore innovative ways of interdisciplinary research in order to improve the understanding of the normative implications of artificial intelligence (AI).
The workshop was divided into two case studies. The case studies highlighted various aspects and concerns raised in the public debate on AI, such as discrimination, data collection, surveillance and the commercial potential of the technology. By analysing two cases, the researchers with different backgrounds were able to find a common basis and successfully lay the foundation for future discussions in research and identify governance issues. UZH News and More...
The workshop was organized by the Chair of Sociology of Law with special focus on media law (Prof. Christoph B. Graber) at the UZH Faculty of Law in cooperation with the Berkman Klein Center for Internet and Society at Harvard University. The workshop was financially supported by the Faculty of Law and the DSI.