HCDS at EACL 2026 – A Strong Presence in Multilingual & Inclusive AI

23 February 2026, by Janis-Marie Paul

Photo: Janis-Marie Paul/UHH

The Hub of Computing and Data Science (HCDS) will be represented at EACL 2026 (European Chapter of the Association for Computational Linguistics) with a diverse and impactful portfolio of accepted papers. The contributions span multilingual NLP, low-resource language technologies, retrieval-augmented generation (RAG), multimodal reasoning, healthcare AI, and large-scale community analysis within AfricaNLP research.

These works reflect HCDS’s core research priorities: inclusive AI, robust language technologies for underrepresented languages, data-driven methods for societally relevant challenges, and interdisciplinary collaboration across computer science, linguistics, and applied domains.

Main Trac

• T2-RAGBench: Text-and-Table Aware Retrieval-Augmented Generation
  Jan Strich, Enes Kutay Isgorur, Maximilian Trescher, Chris Biemann, Martin Semmann
  T2-RAGBench introduces a new benchmark for evaluating retrieval-augmented generation systems that integrate reasoning over both structured (tables) and unstructured (text) data. The work addresses a critical challenge in current RAG systems: consistent reasoning across heterogeneous information sources.

Student Research Workshop (SRW)

We are particularly proud of two papers supervised by Jan Strich that were accepted to the Student Research Workshop:

• Comprehensive Comparison of RAG Methods Across Multi-Domain Conversational QA
  A systematic comparison of RAG approaches in multi-domain conversational question answering.
  Klejda Alushi, Jan Strich, Chris Biemann, Martin Semmann
  
  
• LEMUR: Robust Fine-Tuning for Multilingual Embedding Models for Retrieval
  Exploring robust fine-tuning strategies for multilingual embedding models in retrieval settings.
  Narges Baba Ahmadi, Jan Strich, Martin Semmann, Chris Biemann

AfricaNLP Workshop

HCDS contributes multiple papers to the AfricaNLP Workshop, focusing on African languages, healthcare accessibility, and socially relevant NLP:

• LexiMCH: A Bilingual Medical Knowledge Lexicon for Maternal and Child Healthcare in Low-Resource Settings
  Structured bilingual medical knowledge resources to support healthcare accessibility.
  Aziza Umer Yibrie et al. (Seid Muhie Yimam and Katrin Schöning Stierand)
  
  
• Can I Read My X-Ray Report? Towards Accessible Radiology Reports in Low-Resource African Contexts
  Improving patient-centered access to radiology information.
  Aziza Umer Yibrie et al. (Seid Muhie Yimam)
  
  
• Media Framing Analysis of Ethiopian Conflict: Combining MAXQDA and NLP
  Bridging qualitative analysis and computational methods for conflict research.
  Adem Chanie Ali & Seid Muhie Yimam
  
  
• Full Fine-Tuning vs. Parameter-Efficient Adaptation for Low-Resource African ASR
  A controlled study on adapting Whisper models for African speech recognition.
  Sukairaj Hafiz Imam et al. (Seid Muhie Yimam)
  
  
• AmharicStoryQA: A Multicultural Story Question Answering Benchmark in Amharic
  A new benchmark strengthening QA research in African languages.
  Israel Abebe Azime et al. (Seid Muhie Yimam)
  
  
• Trust but Check: LLM-Assisted Review of Human Translations in African Languages
  Exploring collaborative human–LLM workflows for translation quality assurance.
  Tadesse Destaw Belay et al. (Seid Muhie Yimam)
  
  
• The Rise of AfricaNLP (2005–2025)
  A large-scale analysis of AfricaNLP’s growth and community impact.
  Tadesse Destaw Belay et al. (Seid Muhie Yimam)

SIGHUM Workshop (LaTeCH-CLfL) 2026

• Narrative in Short German Prose: A Multi-Phenomenon Dataset for Computational Literary Analysis
  A novel dataset for computational literary analysis of narrative phenomena in German prose.
  Hans Ole Hatzel, Haimo Stiemer, Evelyn Gius, Chris Biemann

With this strong presence, HCDS reinforces its role as an interdisciplinary hub for multilingual, inclusive, and socially impactful AI research.

We look forward to engaging with the community at EACL 2026.