Profile

I am a Professor for Data Science at Universität Koblenz and Visiting Professor at the University of Sheffield. In Koblenz, I lead the Institute for Web Science & Technologies and coordinate the MSc in Web and Data Science and BSc in Computational Social Science programmes.

My main research interest is in the intersection of human-centred artificial intelligence, data science, and computational social science. In particular, I am interested in understanding and modelling how data-driven systems (e.g., information retrieval and recommender systems) shape and influence people and society. Often, I collaborate with partners from industry or other disciplines interested in developing personalisation services or in gaining new insights into their data or society.
My work is underpinned by data mining, natural language processing, machine learning, and human-computer interaction techniques.

I am an active member of the research community. I have hold roles such as steering committee member of the ECIR and ACM SIGIR ICTIR conferences, general chair of ACM CIKM'23 and ACM CHIIR'24, or was co-organiser of evaluation campaigns at CLEF, NTCIR, and MediaEval. I am on the editorial board of the IP&M journal and was guest editor for various special issues of other journals.

I have completed my PhD in Computing Science at University of Glasgow under the supervision of Prof. Joemon M. Jose and Prof. Keith van Rijsbergen. After this, I have worked at research-led science & engineering centres in the US and Ireland, as well as at universities in Germany, China, and the UK. Since 2022, I am a Senior Fellow of the (UK) Higher Education Academy.

Research

Data Science

My team deals with all aspects of the acquisition, management and analysis of data and information with the aim of extracting novel insights. In particular, we are interested in exploring how to generate insights from personal data and user-generated content. Often, the aim of the analysis of such data is to gain valuable insights into users’ opinions, interests, and behaviour, or to gain new insights into society.

• Personal data relates to any information that a living person can be identified from. Example data that we have worked with in the past include healthcare data, digital cultural heritage records, users’ online transaction logs, or personal records generated by Wearable devices or self-tracking apps, referred to as Lifelogs. We are also interested in data organisation and privacy-aware data processing (e.g., in the form of Evaluation-as-a-Service).

• User-generated content refers to text or audio-visual content that has been generated by users. For example, we have published work on the analysis of social media posts (e.g., for hate speech classification) and e-mails (to visualise communication patterns), and conducted a forensic analysis of files stored on individuals’ hard drives (to reconstruct creative working patterns).

Human-centred Artificial Intelligence

We study how artificial intelligence techniques can be employed to create more human-centred online experiences. We mainly concentrate on interactive Web systems (i.e., information retrieval and recommender systems) that assist users in coping with information overload caused by the availability of large amounts of data and information.

• Information retrieval (IR) systems provide access to information relevant to their information need, usually expressed in the form of search queries. Web search engines are the most popular IR applications but they can also be found in more domain-specific scenarios, e.g., in enterprise settings, archives, or libraries. Our main contributions are in the fields of user modelling, interactive information retrieval, and in enhancing the search experience (e.g., via gamification).

• Recommender Systems are information filtering systems that seek to predict user preferences for a set of items (such as books, movies, music, etc.). Recommender systems can often be found in the e-commerce sector where they are employed to recommend products to customers. We have contributed to advances in different recommendation domains (e.g., news or television recommendation) and stream-based recommendation. Further, we have promoted the first living lab for the evaluation of information access systems to support large-scale benchmarking.

Computational Social Science

We are interested in exploring societal challenges caused by the rise of data science and artificial intelligence. In the context of data-driven systems, we are studying issues related to fairness, accountability, transparency, and ethics (FATE). We also focus on researching approaches for for socially responsible data science education.

• Fairness, accountability, transparency, and ethics can be seen as one of the grand challenges of data-driven systems. Prior research has shown that many algorithms that underlie modern personalisation systems are not transparent or neutral. They carry social values; some exhibit biases or systematically produce results that could lead to discrimination against certain people. My team is interested in studying the impact of information bias and in shedding light on the need for algorithmic transparency.

• Socially responsible data science education involves raising people’s critical awareness of the power dynamics and potential social consequences of applied AI. We have made contributions in the development of a curriculum for human-centred data science programmes and studied issues related to bridging the digital divide in higher education.

Research Projects

My research projects are often interdisciplinary in nature to address research questions from different angles. Most recently, I have collaborated with professionals from the following sectors:

• Healthcare (e.g., to predict adverse outcome of COVID-19 cases)

• GLAM (Galleries, Libraries, Archives, Museums) (e.g., to process personal digital archives)

• Information and Data Service Providers (e.g., to research personalised recommendation techniques)

• Telecommunications (e.g., for predicting customer churn)

• Arts and Humanities (e.g., to ease access to historic documents)

Supervision

I take an active role in supervising the research efforts of my team, with the aim of enabling them to publish and present their work at leading academic conferences and journals. 

Publications

Most of my publications can be downloaded from open access repositories. Older publications are archived in repositories of The University of Sheffield, The University of Glasgow, TU Berlin, Dublin City University, and of UC Berkeley.

Other publication lists can also be found on ORCID and on Google Scholar.

Service

I actively engage with the research community, e.g., by organising conferences, reviewing and assessing research contributions, or by building bridges between academics and practitioners.

Conference-related activities

Workshops, tutorials, and evaluation campaigns

Outreach and public engagement

Editorials and reviewing

Assessment and Examination

Talks

I am frequently invited to present my research to different audiences. This includes talks to experts from various subject areas but also to the general public. I adapt the focus and language of my presentations based on the interests and prior knowledge of the audience.

To date, a majority of my presentations has been targeting technical audiences. For instance, I have given guest lectures or seminar talks at computer science departments or companies, as well as keynote presentations and invited talks at technical conferences and workshops, most of which were co-located with conferences sponsored by special interest groups of the ACM.

I have also presented to the general public and to academics and practitioners of other subject areas, e.g., at public lectures or at seminar talks organised by departments of digital humanities, philosophy, business, health science, or library, archives and information science.