Hiring: Student Assistant (HiWi) for SpatialBenchRAG Project

We are looking for a motivated student assistant (HiWi) to join the SpatialBenchRAG project, starting November 2025 for a period of 6 months (36 hours/month).

About the Project

The SpatialBenchRAG project addresses the lack of standardized resources for spatial reasoning. The goal is to develop the first multimodal, multilingual benchmark for evaluating retrieval-augmented generation (RAG) systems grounded in spatial knowledge graphs (KGs). The benchmark will combine diverse geographic datasets, including OpenStreetMap and Wikidata, to support realistic, context-aware evaluation across tasks such as spatial question answering (QA), multi-hop reasoning, and visual queries. For more information, visit the project webpage.

Your Role

As a student assistant, you will actively contribute to the development of the benchmark and will be involved in:

• Integrating diverse data formats (text, coordinates, images) from multiple sources (e.g., Wikidata, OpenStreetMap)
• Automatic question-answer generation using Large Language Models, including quality evaluation of outputs
• Contributing to machine translation of datasets and post-editing annotation guidelines
• Evaluation of state-of-the-art models (LLMs, VLMs, RAG approaches) on tasks such as visual QA and spatio-temporal reasoning
• Contributing to scientific publications emerging from the project

Requirements

We are looking for a student who is eager to learn, explore, and contribute to open science research. You should:

• Be enrolled in the Master of Business Informatics or Data Science at a German university
• Have completed a Bachelor degree
• Have strong programming skills in Python; experience with PyTorch and HuggingFace Transformers is a plus
• Have completed some of the following courses:
  • Advanced Methods in Text Analytics (IE 696)
  • Information Retrieval and Web Search (IE 663)
  • Knowledge Graphs (IE 650)
  • Text Analytics (IE 661)
• Be interested in RAG, NLP, or spatial data applications

Why Apply?

This position offers a unique opportunity to:

• Gain hands-on experience with diverse multimodal data (text, spatial, visual)
• Work with state-of-the-art Large Language Models and RAG methods
• Explore multiple research directions (NLP, RAG, KG, multimodality)
• Co-author publications at top-tier conferences
• Contribute to open science and reproducible research

Application Details

Please send the following documents in a single PDF to andreea.iana (at) uni-mannheim.de:

• Curriculum Vitae (CV)
• Transcript of records (Bachelor’s + current Master’s transcript)
• If possible, a copy of your Bachelor thesis

Applications will be considered on a rolling basis until the position is filled.

If you are passionate about RAG, multilingual / multimodal NLP, or spatial data, we strongly encourage you to apply!