Magdalena Kaiser is a PhD researcher in the Databases and Information Systems Group at the Max Planck Institute for Informatics (MPII), Saarbrücken, Germany, under the supervision of Prof. Gerhard Weikum and Dr. Rishiraj Saha Roy. Her research focuses on conversational question answering that require understanding of the conversational context to answer incomplete follow-up user questions. In particular, the goal of her thesis is to improve conversational systems by learning from limited amount of data and implcit feedback using Reinforcement Learning. In her work, she applies techniques from Machine Learning, Natural Language Processing and Information Retrieval. Prior to joining MPII, she worked as a researcher at the German Research Center for Artificial Intelligence (DFKI). She obtained her Master's degree from Saarland University and her Bachelor's degree from University Erlangen-Nürnberg, Germany.

• (Conversational) Question Answering
• Machine Learning (in particular Reinforcement Learning)
• Knowledge Graphs

• Magdalena Kaiser and Gerhard Weikum, Preference-based Learning with Retrieval Augmented Generation for Conversational Question Answering, WWW 2025. [Preprint] [Website] [Code] [Poster]
• Magdalena Kaiser, Patrick Ernst and Gyuri Szarvas, Learning from Relevant Subgoals in Successful Dialogs using Iterative Training for Task-oriented Dialog Systems , EMNLP 2024. [Preprint].
• Magdalena Kaiser, Rishiraj Saha Roy and Gerhard Weikum, Robust Training for Conversational Question Answering Models with Reinforced Reformulation Generation, WSDM 2024. [Preprint] [Website] [Code]
• Magdalena Kaiser, Rishiraj Saha Roy and Gerhard Weikum, Reinforcement Learning from Reformulations in Conversational Question Answering over Knowledge Graphs, SIGIR '21. [Preprint] [Data+Demo] [Code] [Slides] [Poster] [ACM Badge]
• Magdalena Kaiser, Incorporating User Feedback in Conversational Question Answering over Heterogeneous Web Sources, SIGIR 2020.
• Magdalena Kaiser, Rishiraj Saha Roy and Gerhard Weikum, Conversational Question Answering over Passages by Leveraging Word Proximity Networks, SIGIR 2020. [Preprint] [Demo]
• Magdalena Kaiser, Rishiraj Saha Roy and Gerhard Weikum, CROWN: Conversational Passage Ranking by Reasoning over Word Networks, 28th Text Retrieval Conference (TREC), 2019. [Preprint] [Code] [Slides] [Poster].
• Magdalena Kaiser and Christian Bürckert, Context-Based Multimodal Output for Human-Robot Collaboration, 11th International IEEE Conference on Human System Interaction (HSI), 2018.

Programm committee member:

• SIGIR 2025 (Demo Track)
• ACL 2023
• SIGIR 2023 (Demo Track)
• SIGIR 2022 (Demo Track)
• WSDM 2022
• AAAI 2022

• Teaching Assistant for Selected Topics in Question Answering (Seminar, 7 ECTS) at Saarland University, Winter Term 2020/2021
• Teaching Assistant for Question Answering Systems (Advanced Lecture, 6 ECTS) at Saarland University, Summer Term 2020
• Teaching Assistant for Information Retrieval and Data Mining (Core Lecture, 9 ECTS) at Saarland University, Winter Term 2019/2020
• Bachelor Thesis supervision of Adina Pohle, Title: Teaching Object Properties to a Robot Multimodally at Saarland University (10/2018-10/2019)
• Teaching Assistant for AI for the Transfer of Control from Autonomous Systems to Humans (Seminar, 6 ECTS) at Saarland University, Summer Term 2018

• September 2023 - February 2024: Applied Science Intern at Amazon, Berlin
• July 2017 - June 2019: Researcher at the Cognitive Assistants Department at the German Research Center for Artificial Intelligence (DFKI), Saarbrücken
• October 2015 - December 2016: Research Assistant at the Cognitive Assistants Department at the German Research Center for Artificial Intelligence (DFKI), Saarbrücken
• October 2012 - September 2014: Working Student at Siemens AG, Nürnberg

• July 2019 - present: PhD student in Computer Science at Saarland University and the Max Planck Institute for Informatics (supervisors: Prof. Gerhard Weikum and Dr. Rishiraj Saha Roy)
• October 2014 - May 2017: Master in Computer Science at Saarland University, Title of Master's Thesis: Multimodal Fission with a Focus on Collaborative Human-Robot Interaction (supervisor: Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster)
• October 2011 - September 2014: Bachelor in Computer Science at Friedrich-Alexander University Erlangen-Nürnberg, Title of Bachelor's Thesis: Combinatorial Auction for Resource Allocation on Multicore Processors (supervisor: Prof. Dr.-Ing. Jürgen Teich)

• Google Scholar
• DBLP
• GitHub
• LinkedIn