Program

We are excited to present the draft program for CONVERSATIONS 2025. Two days brimming full of sharing and discussion on research on chatbots and human-centred AI (November 12-13). All times in Central European Time.

All presentations will be held on-site to facilitate networking and interaction. All plenary sessions will be available for an online audience for broadened engagement and participation (hybrid mode).

Paper presentation in plenary sessions: 8-10 minutes presentation time, 5-7 minutes of Q&A.

DAY 1 (November 12, 0930-1730)

0900-0930. REGISTRATION
0930-0950. OPENING
  • Welcome speech. Prof. Dr.-Ing. Jochen Abke, Vice President for Studies and Digitalization, TU Lübeck
  • Introduction to the day. Asbjørn Følstad
0950-1050. PLENARY SESSION 1: INTERACTIONS AND RELATIONSHIPS

Chair – Asbjørn Følstad, SINTEF

  • Barriers shaping synthetic relationships with companion bots: A three week initial adaptation study (full paper). Marisa Tschopp, Nadine Schlicker, Nima Zargham, Henrik Skaug Saetra and Johannes Schöning.
  • Evaluating the usability of a social robot for book talks in primary school (full paper). Mattias Arvola, Linda Söderqvist, Emma Mainza Chilufya, Anna Martín Bylund, Susanne Severinsson, Linnéa Stenliden and Tom Ziemke
  • Student counsellors’ perspectives on a student mental well-being chatbot: Exploring needs, preferences, and design requirements (full paper). Valentina Bartali, Emmelyn A. J. Croes, Tibor Bosse, Renate H. M. de Groot and Marjolijn M. L. Antheunis
  • Left, right, down, or round – Exploring the design of patterns in chatbot interfaces and how users like them (full paper). Verena Traubinger, Sebastian Heil, Akshata Anil Mithari and Martin Gaedke

1050-1120. COFFEE BREAK

1120-1245. GROUPWORK – HOW LARGE LANGUAGE MODELS IMPACT SCIENTIFIC RESEARCH

Large language models already impact scientific research – for better or worse – and the impact is arguably likely to grow. As researchers and practitioners with an interest in chatbots and human-centred AI, we are well positioned to analyse and inform the way LLMs are taken up for research purposes. In this groupwork you are invited to share practical examples, reflect on the path forward, and discuss how we as a research community should take on LLMs going forward.

1245-1345. LUNCH
1345-1430. POSTER SESSION 1

Poster authors will be available for presentation and discussion by their posters. This is a great opportunity for symposium participants to engage with the presented research.

  • Seeing you seeing me: Augmenting human-robot dialogue with vision language models. Thomas Sievers
  • Support style-oriented in-context learning example selection for empathetic response generation during customer support conversations. W L Yeung
  • Generative artificial intelligence in qualitative research methods: Between hype and risks? Maria Couto Teixeira, Marisa Tschopp and Anna Jobin
  • The AI2Entrepreneur Promptathon: Fostering AI literacy and entrepreneurship education through interactive workshop formats. Miriam Maibaum and Benedikt Breitkopf
  • Relationship formation with artificial companions. Ezgi Dede, Hande Sungur, Jeroen Lemmens and Jochen Peter
  • Mapping and calibrating user trust with LLMs: First steps towards developing a framework for shaping trust. Samuel Hill, Joy Belgassem and Felix Nadolni
  • Across the domains: Investigating interface and pragmatic features. Johanna Rockstroh, Laura Spillner, Nima Zargham, Rainer Malaka and Nina Wenig
  • Evaluating AI-authorship: The role of text domain and presentation style. Angelica Henestrosa, Julia Schühle, Joachim Kimmerle and Teresa Luther
  • Design and evaluation of Eva.ai: A conversational AI learning environment for programming education. Paraskevas Lagakis and Stavros Demetriadis
  • From voice to feeling: Lessons learned from a hybrid deep-learning model for spanish emotion recognition in virtual assistants. Rafael del Hoyo and Gorka Labata
1430-1515. KEYNOTE: PROF. DR. JULIAN KUNKEL

In his keynote talk, Prof. Dr. Julian Kunkel, GWDG – Gesellschaft für Wissenschaftliche Datenverarbeitung mbH Göttingen (and will share from the building the large scale KISSKI AI platform (https://gwdg.de/en/projects/kisski-overview/kisski/) – mainly based on open source models (LLMs and image processing) – and offering this to universities and institutions. (read more about the keynote talk)

1515-1600. COFFEE BREAK
1600-1700. PLENARY SESSION 2: HUMAN-CENTRED AND TRUSTWORTHY

Chair: Sebastian Hobert, Technische Hochschule Lübeck

  • Human or humane? Users’ pluralist perspectives on the ethical values and risks of digital humans (full paper). Lotte Willemsen, Iris Withuis, Sanne Snoeij-Smit and Charlotte van Hooijdonk
  • Balancing offensive language and freedom of expression in AI speech assistants: An ethical and algorithmic perspective (full paper). Madeleine Rischer, Jonas Ehrhardt and Gerhard Schreiber
  • Measuring political chatbots: A multi-dimensional framework for trustworthiness (full paper). Thilo Dieing
  • FairBotBench: A bilingual benchmark for evaluating chatbot behavior in e-commerce (full paper). Maik Thiele, Jürgen Anke, Steffen Tomschke, Tamás Janusko, Ricardo Bochnia and Gunnar Hempel
1700-1715. LIGHTNING TALKS

Participants are invited to sign up for 2-3-minute presentations on a topic they want to promote – e.g. ongoing research, upcoming events, or other themes of interest. Sign up by contacting the coordinators – by email or in person during the symposium.

1715-1730. WRAP UP DAY 1 AND BEST PAPER AWARD CANDIDATES
2000 –   SOCIAL DINNER (self cost)

DAY 2 (November 13, 0900-1500)

0900-0915. WELCOME
0915-1045. PLENARY SESSION 3: CO-DESIGN AND USER EXPERIENCE

Chair: Effie L.-C. Law, Durham University

  • Social robot agency and fiction: Co-designing a conversational robot for reading with children (full paper). Linnéa Stenliden, Anna Martín Bylund, Emma Mainza Chilufya, Susanne Severinsson, Mattias Arvola and Tom Ziemke
  • Rewriting the script: Conversation design’s evolution with LLM co-design (full paper). Melissa Guyre, Lisette Gonzalez, Meryll Davis and Rahul Divekar
  • “It’s mysterious, bodiless and omniscient, like a god”: Exploring users’ perceptions of synthetic relationships in a mental health app (full paper). Maria Couto Teixeira, Annika Aebli, Stefanie Klein and Marisa Tschopp
  • Older adults’ perceptions of LLM-based music chatbots (full paper). Farkhandah Aziz Komal, Effie Lai-Chong Law and Martha Correa-Delval
1045-1115. COFFEE BREAK
1115-1200. POSTER SESSION 2

Poster authors will be available for presentation and discussion by their posters. Symposium participants are warmly welcome to engage with the presented research.

  • Towards an open platform for evaluating conversational assistants with empathy, negotiation and procedural fairness. Rafael del Hoyo-Alonso, Patricia Pérez-Curiel, Rosa María Montañés and Juan Carlos Bustamante
  • Care-AI – paving the way for AI in elder care. Rebecca Starke, Christina Bober, Kristina D’Alessio, Eileen Heydenreich and Prof. Dr. Beatrice Podtschaske
  • Adaptive scaffolding for human-centered AI in education. Giuliana Lavagnino and Michael Wessel
  • MediCall: A voice-based booking system for non-emergency medical transports. Hannes Bradl, Ferdinand Fuhrmann, Melissa Hagendorn, Marlene Sarka and Franz Graf
  • FairBotLive: Exploring bias, personalization, and toxicity in LLM-powered chatbots. Maik Thiele, Jürgen Anke, Ricardo Bochnia, Steffen Tomschke, Tamás Janusko, Tanja Dietrich, Paulus Tamm and Gunnar Hempel
  • Designing usable interfaces for human evaluation of LLM-generated texts: UX challenges and solutions. Androniki Mertsiotaki, Stephanie Hofmann, Sarah Keck, Alexander Daum, Emily Kratsch and Birgit Popp
  • Hybrid (rule-based + gen AI) chatbot: Automating negotiations. Klaus Garrido Tenorio, Michael Becker-Peth and Jelle de Vries
  • Generative AI and IT professionals:  A study of usage practices. Anne-Sofie Færk Spens, Kristian Grønlund Mikkelsen, Tanja Svarre and Marianne Lykke
1200-1230. GROUPWORK – PRESENTATIONS FROM BREAKOUT SESSIONS

Each group presents outcomes of the groupwork. Presentation time 2-3 minutes pr. group, with additional time for questions and discussion.

1230-1330. LUNCH
1330-1445. PLENARY SESSION 4: APPLICATIONS FOR PROFESSIONAL CONTEXTS

Chair: Asbjørn Følstad, SINTEF

  • LLM usage for clinical decision support and documentation support in hospitals – a systematic review (full paper). Laura Anderle, Abdelbaset Abidi, Saban Ünlü, Oliver Steidle, Tim Janßen and Thomas Petzold
  • A hybrid semantic-lexical FAQ chatbot architecture with paraphrase augmentation and generative response refinement (full paper). Mazen Salem Asag, Mueeze Mushabbir, Kamrul Hasan and Hasan Mahmud
  • Motion2Meaning: A clinician-centered framework for contestable LLM in parkinson’s disease gait interpretation (full paper). Phuc Truong Loc Nguyen, Thanh Hung Do, Truong Thanh Hung Nguyen and Hung Cao
  • Generative AI in the workplace: workers’ experiences and competence development (full paper). Anna Grøndahl Larsen, Marita Skjuve and Asbjørn Følstad
  • Not all trust is the same: Effects of decision workflow and explanations in human-AI decision making (full paper). Laura Spillner, Rachel Ringe, Robert Porzel and Rainer Malaka

1445-1500. WRAP-UP AND BEST PAPER AWARD