PhD on contextualizing generative AI systems in structured data (m/f/x)

Goal of the DataLibra project
Approximately 120 zettabytes of data has been collected worldwide but less than 1% is actually
used. Structured data, e.g. tables, spreadsheets, and relational databases, is prevailing in
organizations and typically informs important decisions in healthcare, governments and finance.
Yet, while AI has demonstrated a high impact on applications on text and images, proportional
progress on structured data is lacking. With the DataLibra project, we aim to close this gap, by
developing AI models and tools for structured data (Table Representation Learning), to help
organizations, of any size, domain, and level of data literacy, get insights from structured data,
efficiently, accurately and securely.

Goal of this PhD project
Following recent developments in AI, large language models (LLMs) have been explored for data
analytics tasks (e.g. text-to-sql), but show limited accuracy in domain-specific contexts with
structured data. In this project, we will investigate and design interactive intelligent systems for
data analytics tasks, while accounting for two key challenges: trustworthiness of the outputs
(factuality), and security constraints of proprietary data contexts as in healthcare, enterprises, and
governments. Potential directions to explore are agentic systems, retrieval augmented generation,
(instruction) fine-tuning, and others.

What you will be doing
 Inform a research agenda on the PhD topic for a timespan of four years.
 Develop methods and systems for contextualizing generative AI for analytics over
structured data.
 Actively collaborate with other researchers in the DataLibra project (students, 4-5 PhDs,
postdocs, PI) and external collaborators (e.g. Amsterdam UMC, University of Amsterdam).
 Communicate research outcomes through papers and presentations at conferences,
workshops and other (scientific) gatherings.
 Assist in relevant teaching activities at universities, such as thesis supervision and assisting
in courses.

Postdoc Trustworthy Human-AI interaction for Media and Democracy (m/f/x)

The AI, Media and Democracy (AIMD) ELSA Lab’s aim is to set up an experimental space and testbed for novel ways of applying AI in the media. Journalists, media professionals, designers, citizens, researchers and public and societal partners bring in their challenges in using AI in the media. The lab investigates how AI-driven applications in the media change the public sphere, civic engagement and economic competition, and explores novel ways of applying AI in the media. The Distributed & Interactive Systems (DIS) research group at CWI combines data science with a strong human-centric, empirical approach to understand the experience of users. This enables us to design and develop next generation intelligent and empathic systems. We base our results on realistic testing grounds and data sets, and embrace areas such as ubiquitous and affective computing, human-centered multimedia systems, and human-AI interaction.

Encounters with AI-generated content can impact the human experience of algorithms, and more broadly the psychology of Human–AI interaction. In particular, AI system disclosures can influence users’ perceptions of media content. There is a growing concern that as generative AI becomes more widely used, manipulated content could easily spread false information. As a key step toward mitigating harms and risk, a key recent mitigation effort involves the European Al Act Proposition “Transparency Obligations for Providers and Users of Certain AI Systems and GPAI Models”, which seeks to address the issue of AI system transparency.

The research scope broadly addresses the effective, trustworthy, and transparent communication of AI system disclosures. We aim to account for ethical and legal considerations, design and human factors perspectives, as well as policy recommendations. As such, this role may involve relevant stakeholders where necessary, ranging from media organizations, policy makers, as well as AI researchers and practitioners. The initial focus is on the end-user (media consumer) perspective, and at later stages, on the perspective of the media organizations and the generative AI media production process itself. For this postdoc, we are specifically interested in how trust in AI systems can be garnered by focusing on creating better user interfaces and/or understanding human-AI system interactions at a cognitive, behavioral, and physiological level. By establishing user-centric designs for transparent AI disclosures, we can take steps toward ensuring a well-functioning democratic society.

We expect the postdoc researcher to be embedded within the AI, Media, and Democracy lab, which is located in Amsterdam’s city center.

PhD student on the subject of Responsible AI for Cultural Heritage (m/f/x)

Are you inspired by human-centered, inclusive AI? And do you have an affinity with cultural heritage? In the HAICu project, we investigate AI for and with the cultural heritage sector, to make cultural heritage (CH) data accessible to citizens, journalists, and other stakeholders. HAICu actively involves users, so that they can be more than passive recipients of information and can help to describe and enrich CH data. Thus, the HAICu project aims to generate and utilize three types of complementary metadata — traditional metadata, citizen-contributed descriptions, and AI-generated labels. This approach ensures a breadth and variety of perspectives. However, each type of metadata can have its own biases. In this project, we are focussed on identifying and addressing the different perspectives and biases in each source. We tackle these issues in both fundamental and applied research in collaboration with the National Museum of World Cultures. In this context, we are looking for a talented PhD student to study one or more of the following topics:

• Bias in AI: methods and metrics to determine to what extent an AI pipeline propagates, exacerbates or introduces bias;
• Inclusivity in AI: methods and guidelines to increase fairness and diversity in data and systems.
• Linked Open Data: explicit representations of provenance and perspectives in data; investigations into bias in Linked Open Data.
• Human computation and citizen science: integration of citizen-contributed labels for cultural heritage objects with professional metadata and AI-generated labels; investigating bias, interrater agreement and quality of such citizen-contributed labels.

Within this broad range of topics, student and supervisor will together decide on specific research questions, taking into account the expertise and interests of the student, the needs of the HAICu project, and the latest developments in the academic field of Responsible AI.

HAICu (digital Humanities - Artificial Intelligence - Cultural heritage) is a large-scale Dutch research project in which AI researchers, Digital Humanities scholars and a wide range of cultural-heritage institutions collaborate. Within the HAICu team, the PhD researcher will participate in Work Package 5, titled “Construction of polyvocal, multimodal narratives.” In this Work Package, CWI will collaborate with UvA and VU and with the National Museum of World Cultures.

The researcher will be based in a dynamic research group called Human-Centered Data Analytics (HCDA). HCDA investigates human-centered, responsible AI in the culture and media sectors. How can we ensure that digital systems are inclusive, promote diversity and can be used to combat misinformation? The HCDA group addresses these important questions. Our work includes a wide range of techniques, such as statistical AI (machine learning), symbolic AI (knowledge graphs, reasoning), and human computation (crowdsourcing). By analyzing empirical evidence of human interactions with data and systems, we derive insights into the impact of design and implementation choices on users. We maintain close collaborations with professionals from the culture and media sectors, as well as social scientists and humanities scholars, through the Cultural AI Lab and the AI, Media and Democracy Lab. These interdisciplinary labs provide us with opportunities to work with real data and real-world use cases.

Two PhD positions in AI-based methods for integrated hydrogen (m/f/x)

[1] The first position is part of WP2 of GVNL, that covers the transport and storage of hydrogen. The PhD position at CWI will concern the interface between hydrogen and electrical power network, in particular how AI optimisation and modeling methods for hydrogen storage can be used/operated in a smart way to provide flexibility for the electrical power network (such as how hydrogen units can act as a form of energy storage for renewable generation). Potential topics for this PhD position include:

• Modelling the deployment of hydrogen storage for managing electricity network congestion, including optimizing their spatial placement and size
• New AI/ML algorithms for smart control of H2 generation/storage units (e.g. reversible H2 fuel cells)
• Novel methods for market-based coordination, and automated market design (using smart autonomous agents) when using H2 assets as a source for flexibility for local power grids

[2] The second position is part of WP3 of GVNL, that concerns end uses of hydrogen, including in local applications such as industrial parks, farms and the transport network (including hydrogen-based trucks and inland shipping). The PhD position at CWI will research how AI and multi-agent techniques can be used to model and optimise complex integrated energy systems that include hydrogen, especially in transportation chains and the energy sector. Specific potential topics include:

• Algorithms for optimising the placement of hydrogen refueling stations, including freight and ship bunkering (refueling) stations
• Models of local flexibility in microgrid applications that include hydrogen (farms, industrial parks etc)
• New automated, agent-based hydrogen markets, especially related to transportation applications and local energy microgrids

The focus of both PhD projects at CWI, will be on development of novel dedicated approaches with (distributed) AI, multi-agent systems and/or machine learning techniques to support decision-making for integrated hydrogen-electricity systems. Fundamental AI techniques relevant to this project include a wide range of computational intelligence and ML techniques, distributed, multi-agent optimisation, market-based coordination between agents, automated negotiation and algorithmic game theory approaches. In terms of the optimization aspects, relevant topics include: single and multi-criteria optimization, dynamic constraints, uncertainty and risks, costs of actions, and economic/technical trade-offs. The PhD student will start with a survey identifying promising alternatives and then delve deeper into specific solutions and approaches and real case studies and data. The projects will involve regular discussion and presentation of the results to other consortium members from other universities, HBOs, research institutes and companies (who have, potentially, different expertise backgrounds). Hence effective communication skills are required

The PhD students will be employed for 4 years in the Intelligent and Autonomous Systems Group (IAS, https://www.cwi.nl/en/groups/intelligent-and-autonomous-systems/) at CWI, the National Research Institute for Mathematics and Computer Science in the Netherlands, located in Amsterdam. The IAS group has an extensive experience in developing and applying computational intelligent, multi-agent systems and applied game theory techniques to address societal problems, and energy systems are an increasingly important application area. The student with be supervised by two senior researchers in the group: Prof. Valentin Robu (also part-time professor at TU Eindhoven) and Professor Han La Poutré (also part-time professor at TU Delft). While the position is primarily based in Amsterdam, based on a successful dissertation, the students will be awarded the PhD degree from TU Eindhoven..

Projectcontroller (m/v/x) (0,8-1 fte)

met een collega projectcontroller beheer je verschillende projecten. Jij zorgt ervoor dat de projecten op financieel vlak voldoen aan de gestelde subsidievoorwaarden en interne- en externe regelgeving. Je houdt overzicht en haalt deadlines door goed te plannen en prioriteiten te stellen. Hierbij werk je samen met en ondersteun je groeps- en projectleiders zodat zij zich kunnen richten op hun kerntaak: onderzoek. Zo lever jij een bijdrage aan het verantwoord besteden van onderzoeksgeld. Binnen de afdeling Financiële Administratie werken de projectcontrollers nauw samen met de overige medewerkers om zo een betrouwbare (vertaling van de) projectadministratie in de concern systemen te waarborgen.