Kdml May 2024

kdml@cs.uni-kassel.de

4 participants
5 discussions

by Eyke Hüllermeier

Still a few more places available, but registration fee needs to be received by 29 May at the latest. Don't miss this opportunity to learn about Generative AI from the experts. The European Association for Data Science (EuADS) is pleased to announce the 2024 instalment of its popular Summer School series. What: EuADS Summer School on Generative AI When: 18-21 June 2024 Where: Maison d’Accueil (Convent of the Franciscan Sisters), 50 avenue Gaston Diderich, L-1420 Luxembourg-Belair Register until: May 15th, 2024 (extended till booked out) The rise of Generative AI, especially with the advancements in Large Language Models (LLMs), marks a transformative era in artificial intelligence that is expanding across all disciplines. LLMs aim to bridge the communication gap between machines and humans, paving the way for models that can grasp the nuances of human language and generate outputs in various formats that mimic human cognition and creativity. The critical moment for Generative AI came with the adoption of neural networks, particularly transformer-based architectures, which have become its backbone. These models stand out for their profound ability to digest and learn from extensive corpora and datasets, and also to generate original, contextually rich content. But we are just at the beginning. The emerging models present challenges related to ethics, reliability, the way we experiment with these models, the scope of their inferences, their applications to more specific domains, etc. All this has created a vibrant field of work and opens the doors to a community that will find the right forum in this Summer School. The organisers have put together a highly attractive programme addressing the following topics: - Machines That Speak and Imagine: The Role of AI in Audio and Image Generation - LLMs: Foundations, Advancements, and Ethical Considerations - Generative AI for Data Analytics - Robust Evaluation of Generative AI - What Generative AI tells us about ourselves? The EuADS Summer School is preceded by a public event on Tuesday afternoon June 18th during which Peter Flach (University of Bristol, UK) will deliver this year's Sabine Krolak-Schwerdt Lecture entitled 'Data Science in the time of ChatGPT — Why AI isn’t solved, and how data science can help'. The public event will be followed by a welcome reception. The Summer School is planned as an in-person event and the symposium as an hybrid event. For more information including how to register see: https://www.euads.org/fjkdlasjdiglsmdgkcxjhvckh/euads-summer-school-913-487/ For participants there are a limited number of affordable rooms available at the Maison d’Accueil, please have a look at our website. There are several hotels within walking distance. However, please note it is very urgent to make your reservation now because the Councils of ministers of the EU are all held in Luxembourg during the month of June and therefore prices will rise in the coming months. It is also worth nothing that in Luxembourg all public transport is free of charge. We look forward to seeing you in Luxembourg in June.

8 months

37th International Workshop on Description Logics, DL 2024 - Call for Participation

by Andrea Mazzullo

[apologies for multiple postings] DL 2024 CALL FOR PARTICIPATION 37th International Workshop on Description Logics, DL 2024 June 18-21, 2024, Bergen, Norway Website: https://dl2024.w.uib.no IMPORTANT DATES * May 21st, 2024 Registration deadline * June 3, 2024 Camera-ready version due All dates above are ‘Anywhere on Earth’, namely 23:59 UTC-12 * June 18-21, 2024 Workshop It is planned to hold DL2024 as an in-person event, that is, at least one author of every accepted paper has to physically attend the workshop. GENERAL INFORMATION The DL workshop is the major annual event of the description logic research community. It is the forum in which those interested in description logics, both from academia and industry, meet to discuss ideas, share information and compare experiences. The 37th edition will be held in Bergen, Norway, from June 18th to June 21st. INVITED SPEAKERS Camille Bourgaux - CNRS researcher at École Normale Supérieure (part of PSL University), France Gabriele Kern-Isberner - Technical University of Dortmund, Germany Ondřej Kuželka - Czech Technical University in Prague, Czech Republic ORGANIZATION General & Local Chair: Ana Ozaki - University of Oslo & University of Bergen, Norway PC Co-Chairs: Laura Giordano - University of Eastern Piedmont, Italy Jean Christoph Jung - Technical University of Dortmund, Germany Local Sponsorship Chair: Guohui Xiao - University of Bergen, Norway Publicity Co-Chairs: Andrea Mazzullo - University of Trento, Italy Victor Lacerda - University of Bergen, Norway Contact PC co-chairs can be contacted via e-mail: dl2024 “at” easychair.org.

8 months

ACM Transactions on Recommender Systems CFP: RecSys 4 Good

by Marko Tkalcic

CALL FOR PAPERS: ACM Transactions on Recommender Systems Special Issue on Recommender Systems for Good Submission deadline: 1. September 2024 https://dl.acm.org/pb-assets/static_journal_pages/tors/pdf/TORS_SI-Recommen… Guest Editors: - Marko Tkalčič, University of Primorska, Slovenia - Noemi Mauro, University of Turin, Italy - Alan Said, University of Gothenburg, Sweden - Nava Tintarev, University of Maastricht, Netherlands - Antonela Tommasel, ISISTAN, CONICET-UNCPBA, Argentina Recommender systems are among the most widely used applications of machine learning. Since they are so widely used, it is important that we, as practitioners and researchers, think about the impact these systems may have on users, society, and other stakeholders. In practice, the focus is often on systems and values of improving key performance indicators (KPIs), such as increased sales or customer retention. Recommendation technology is currently underutilized to serve societal goals that go beyond the business objectives of individual corporations. However, other values, bound more to societal good, could be considered in the development and goals of a recommender system. In fact, recommender systems have already been explored to stimulate healthier eating behavior and for improved health and well-being in general, to help low-income families make school choices, to suggest successful learning paths for students, to entice climate-protecting energy-saving behavior, to support fair micro-lending, or improve the information diets of news readers. Research in these areas is however limited in numbers, compared to the many papers that are published every year that propose new models for improved movie recommendations. Moreover, concerning the methodology and evaluation perspective in this area, it is essential to find a clear methodology and criteria for evaluating the effectiveness and "goodness" of the proposed algorithms. This includes acknowledging that different values may be conflicting, as well as resolving how and when (and by whom) certain values should be prioritized over others. Research on "Recommender Systems for Good" may benefit from an interdisciplinary approach, drawing on insights from fields such as computer science, ethics, sociology, psychology, law, and economics. Collaborations with stakeholders from diverse backgrounds can enrich the research and ensure that recommendations are grounded in real-world needs and values. This special issue aims to present state-of-the-art research works where recommender systems have a positive societal impact and help us address urgent societal challenges. It will thereby serve as a call to action for more research in these areas. Ultimately, through this special issue, we hope to establish a vision of "Recommender Systems for Good', following the spirit of the "AI for Good" initiative (https://aiforgood.itu.int) to achieve the United Nations Sustainable Development Goals (2015) and the more recent UNESCO recommendation on the Ethics of Artificial Intelligence (2024) (https://www.unesco.org/en/artificial-intelligence/recommendation-ethics). Topics: We aim to collect the latest research on recommender systems for societal good. The topics of the special issues include (but are not limited to): - Recommender systems for safety, security, and privacy (e.g., reducing poverty and inequality) - Recommender systems that protect the environment and ecosystems (e.g., lower energy consumption, water and energy management) - Recommender systems that give control of data back to the users (e.g., transparency of data, models, and outputs) - Recommender systems for the interconnected society (e.g., increase of solidarity, online conversational health, multi-stakeholder recommenders) - Accountability in recommender systems, including addressing emerging regulations, such as the DSA (Digital Service Act) - Recommender systems for the public good (e.g., mental and physical health, welfare, digital literacy, stakeholder engagement, e-learning) - Introspective studies on the current state of RSs concerning societal good - Fairness-preserving and fairness-enhancing recommender systems, unbiased recommendations (e.g. to preserve gender equality) - Responsible recommendation (e.g., in social media and traditional news, avoiding filter bubbles and echo chambers) - Sustainability and Cultural recommendations (e.g., art, cultural heritage) - Recommendations to support disadvantaged groups (e.g., elderly, minorities) - Recommender systems for personal development and well-being (e.g., behavioral change, fitness, self-actualization, personal growth) Important Dates: - Submission deadline: September 1, 2024 - First-round review decisions: December 1, 2024 - Deadline for revision submissions: February 1, 2025 - Notification of final decisions: April 1, 2025 Submissions that are received before the first deadline will be directly sent out for review; papers will be immediately published online after acceptance. Submission Information: The special issue welcomes technical research papers, survey papers, and opinion/reflective papers. Each paper should address one or more of the abovementioned topics or be in other scopes of Recommender Systems for Good. The special issue will also consider peer-reviewed journal versions (at least 30% new content) of top papers from related recommender system conferences such as RecSys, SIGIR, KDD, CIKM, IUI, UMAP, CHI, WSDM, ACL, etc. The new content must be in terms of intellectual contributions, technical experiments, and findings. Submissions must be prepared according to the TORS submission guidelines (https://dl.acm.org/journal/tors/author-guidelines) and must be submitted via Manuscript Central (https://mc.manuscriptcentral.com/tors). For questions and further information, please contact the guest editors at rs4good(a)acm.org. -- ---------------------------------------------------------------------- Dr. Marko Tkalcic http://markotkalcic.com Twitter: https://twitter.com/#!/RecSysMare Linkedin: http://www.linkedin.com/in/markotkalcic Google Scholar: http://scholar.google.com/citations?user=JQ2puysAAAAJ ----------------------------------------------------------------------

8 months

Assistant Professorship AutoML with Tenure Track

by Prof. Dr. Emmanuel Müller

Here's your opportunity to shape the future of Machine Learning and Artificial Intelligence as part of our research institute. The Lamarr Institute [1] and TU Dortmund University [2] are looking for an assistant professorship position with tenure track to be filled as early as possible: Professorship Automated Machine Learning and Optimization (W1 with W2 tenure track) at the Department of Computer Science, TU Dortmund University https://service.tu-dortmund.de/documents/18/2120803/Ausschreibung_W1TTW2_Au… You represent the area of automated machine learning (AutoML) and optimization in research and further develop this with international visibility. The professorship is part of the Lamarr Institute [1] and aligns in research, teaching, and transfer with the five core and five interdisciplinary research areas of the Institute. There, you will establish a new expert group on machine learning equipped with two scientific staff positions for both scientific and application-oriented ML research at TU Dortmund University. In doing so, you participate in a vivid research network and cooperations within and beyond the Lamarr Institute, TU Dortmund University and its AI ecosystem [2]. You are committed to applying for third-party funding as well as promoting early career researchers, and you participate appropriately in the teaching activities of the computer science department. Possible research fields may include but are not limited to: * automated selection of model families * hyperparameter optimization * fine-tuning of models * evaluation of models * automated optimization of neural network topologies * especially in deep learning * reinforcement learning * explainable AI and large language models The research focus can be both on mathematical modelling and the application of AutoML in various domains. If you are interested in the position, please submit your application by May 29th, 2024 via our application portal [3]. Bests, Emmanuel Müller Links: ------ [1] https://lamarr-institute.org/ [2] https://ai.tu-dortmund.de/ [3] https://tudberufung.hr4you.org/job/apply/662/professorship-w1-with-w2-tenur…

8 months, 2 weeks

Postdoc position in Psychologically-informed User Modeling

by Marko Tkalcic

TLDR: postdoc position (PhD required), full-time employment (40 hrs/week), duration 2 years (preferred start October 2024), research and teaching, supervisor: Marko Tkalčič, University of Primorska, Koper, Slovenia Research topics: - psychologically-informed user modeling, - psychologically-informed item modeling, - inference of user and item characteristics, - explanations, - recommender systems. Details: The Department of Information Sciences and Technologies (DIST) at the Faculty of Mathematics, Natural Sciences and Information Technologies (FAMNIT) of the University of Primorska is seeking a top early-career researcher for a postdoctoral position in the area of psychologically-informed user modeling under the supervision of assoc. prof. dr. Marko Tkalčič. The project will explore how psychologically-informed features can be used to model users, items, infer user characteristics from digital traces, infer item characteristics, provide explanations and recommendations in the music and film domains. The methodologies will include user studies and machine learning. The candidate will work in a great team with Marko Tkalčič as part of the HICUP lab (https://hicup.famnit.upr.si/) in the beautiful Mediterranean city of Koper (the beach is only a 3 min walk from the office). - Full-time employment (40 hrs/week). Full social security. - The position is for two years. - The preferred starting date is October 2024. - The candidate must have (or expect to obtain shortly) a PhD in computer science or in an area relevant to the research topics. - The ideal candidate should posses expertise in the following areas: - recommender systems, - machine learning, - user modeling. - Expertise in one or more of the following areas will be appreciated: - ML explainability, - computational psychology, - computational social science, - media-related knowledge (musicology etc.). - The position includes a small teaching load (1 course/semester, 3-4 hrs/week). - The position comes with funding for travel (conferences, visitors). Application Process: The applications will be assessed on a rolling basis until the position is filled. To apply, send an email to marko.tkalcic(a)famnit.upr.si with the following documentation: - motivation letter, - CV, - list of publications, - research statement, - names and email addresses of two to three references. Link to call: https://markotkalcic.com/postdoc_2024.html -- ---------------------------------------------------------------------- Dr. Marko Tkalcic http://markotkalcic.com Twitter: https://twitter.com/#!/RecSysMare Linkedin: http://www.linkedin.com/in/markotkalcic Google Scholar: http://scholar.google.com/citations?user=JQ2puysAAAAJ ----------------------------------------------------------------------

8 months, 2 weeks

2025

2024

2023

2022

2021

Kdml May 2024