Skip to content

Jobs

Student Job — Research Support for a Debate Corpus and AI Experiments

For registered students of the University of Luxembourg exclusively

Position

Official title
Student CDD – Research Support for a Debate Corpus and AI Experiments

Details

  • Expected start: 1 September 2026 or 15 September 2026
  • Working hours: 40 or 60 hours per month
  • Duration: Flexible (minimum 2 months)
  • Location: Belval Campus, Maison du Nombre, University of Luxembourg
  • Application deadline: 27 July 2026 or 7 August 2026 (for 15 September start)

About the Position

The Individual and Collective Reasoning Group is looking for a Bachelor’s or Master’s student to support research on debate, argumentation, and Artificial Intelligence.

The selected candidate will contribute to the construction of a research corpus based on Chinese-language debate shows and will support exploratory Natural Language Processing (NLP) and Large Language Model (LLM) experiments.

Main Tasks

The successful applicant will:

  • prepare, organise, and clean Chinese and/or English transcripts and subtitles;
  • assist in the annotation of debate materials;
  • support exploratory NLP and LLM experiments;
  • document datasets, experiments, and results.

Candidate Profile

Applicants should be enrolled in a Bachelor’s or Master’s programme in:

  • Computer Science
  • Artificial Intelligence
  • or a closely related discipline.

Preferred qualifications:

  • good command of English;
  • Mandarin Chinese proficiency (preferred);
  • attention to detail when working with textual data;
  • basic Python programming skills;
  • interest in debate, argumentation, language technologies, or AI.

Previous experience with NLP, LLMs, corpus construction, or annotation is considered an advantage.

Application

Please submit:

  • Curriculum Vitae (CV)
  • Brief statement of interest

The application procedure will be announced shortly.

Contact

For questions regarding the position, please contact Liuwen Yu.

Master Thesis Topics — CARE-PROOF Collaboration

As part of the CARE-PROOF collaboration with Venture Robot Solutions, we are offering two Master thesis opportunities focused on trustworthy, explainable, and normative behaviour of healthcare service robots.

Both topics build on the existing CareOS telemetry and mission replay infrastructure, with access to relevant robotic platforms, field-deployment data, and technical guidance from Venture Robot Solutions (VRS).

1. Normative Evaluation of Robot Behaviour in Healthcare Environments

Objective:
Develop a systematic and reproducible framework for evaluating whether a service robot behaves appropriately in real-world healthcare environments.

The thesis will investigate how robot behaviour can be evaluated against normative and behavioural expectations, considering situations such as:

  • Maintaining appropriate interpersonal distance
  • Yielding and right-of-way behaviour
  • Waiting and navigation in shared spaces
  • Handling obstacles and unexpected situations
  • Interactions with people in healthcare environments
  • Context-dependent appropriateness of robot actions

The work can build directly on CareOS telemetry, mission replay, and existing field-deployment data to define and validate a set of quantitative behavioural metrics. The resulting framework should enable reproducible evaluation and comparison of robot behaviour across different missions and scenarios.

Key areas:
Robotics · Normative AI · Behaviour Evaluation · Human-Robot Interaction · Healthcare Robotics · Robot Telemetry

2. Explainable Decision Tracing for Healthcare Service Robots

Objective:
Develop a method for reconstructing and explaining why a healthcare service robot made a particular decision during a mission.

The thesis will focus on tracing the robot's decision-making process through the sequence:

Perception → State → Rule → Action

The goal is to transform low-level execution and telemetry data into an interpretable decision trace that explains how the robot moved from its perception of the environment to a particular action.

The work may investigate:

  • Reconstruction of decision-making sequences from CareOS telemetry
  • Linking perceptions and internal states to activated rules
  • Generating human-understandable explanations of robot actions
  • Identifying the reasons behind unexpected or undesirable behaviours
  • Comparing decision traces across different missions and scenarios
  • Supporting auditing and analysis of robot behaviour

The resulting approach should make healthcare robot behaviour more understandable, auditable, and comparable, providing a foundation for explainable and trustworthy robotic systems.

Key areas:
Explainable AI · Robotics · Decision Tracing · Normative AI · Healthcare Robotics · Robot Telemetry

Research Environment

Both thesis topics provide the opportunity to work with real healthcare service robotics data and infrastructure through the CARE-PROOF collaboration with Venture Robot Solutions.

Students will build upon the existing CareOS telemetry and mission replay framework, with access to relevant data and robotic platforms, as well as technical guidance from the VRS team.

The two topics are complementary but sufficiently independent to be pursued as separate Master theses.