Abdelrahman Abdallah

prof_pic.jpg

Digital Science Center (DiSC)

University of Innsbruck

Innsbruck, Austria

Pronouns: He/Him

Hello. Iโ€™m Abdelrahman Abdallah, a Ph.D. graduate in Computer Science advised by Professor Adam Jatowt at the Digital Science Center (DiSC), Department of Computer Science, University of Innsbruck. I graduated with honors in September 2026, receiving Sehr gut (1), the highest grade in the Austrian grading system. Currently, I am a research assistant at DiSC. Previously, I worked as a machine learning engineer at DISCO App in Cairo and at KMG Engineering in Kazakhstan, as a machine learning researcher at Satbayev University and Universitร  Caโ€™ Foscari Venezia, and as a research and teaching assistant at Assiut University in Egypt.

As a researcher in machine learning (ML) and natural language processing (NLP), my work focuses on large language models (LLMs), information retrieval, and question answering. I am particularly interested in how LLMs can be used for document retrieval and reranking, robust QA on noisy or OCR text, and reasoning over temporal information in dynamic collections.

My research interests include:

  • Reasoning-Intensive Information Retrieval & Reranking
  • Efficient & Multimodal Retrieval Models
  • Temporal Information Retrieval & Question Answering
  • Retrieval-Augmented Generation & Agentic Search
  • LLM Evaluation, Calibration & Bias Mitigation

Please feel free to contact me via email at abdelrahman.abdallah@uibk.ac.at or abdoelsayed2016@gmail.com.

news

Sep 28, 2026 ๐ŸŽ“ September 2026 - Graduated with a Ph.D. in Computer Science from the University of Innsbruck with honors, receiving Sehr gut (1), the highest grade in the Austrian grading system.
Aug 22, 2026 ๐ŸŽ‰๐Ÿ“š August 2026 - 10 papers accepted at EMNLP 2026: 9 Main Conference papers and 1 Findings paper.
Aug 21, 2026 ๐Ÿ’ก๐Ÿ› ๏ธ August 2026 - HintEval, an open-source toolkit for hint generation and evaluation, accepted at EMNLP 2026 Demo. ๐Ÿ“„ Paper ยท ๐Ÿ’ป Code.
Aug 13, 2026 August 2026 - Our shared task RETECO: Reasoning-Oriented Retrieval was accepted for SemEval-2027 (GitHub).
Aug 07, 2026 August 2026 - One short paper accepted at CIKM 2026: โ€œDifficulty-Gated Fusion of Reasoning Views for Temporal Retrievalโ€.
May 01, 2026 May 2026 - Two papers accepted: MM-BRIGHT at KDD 2026 and Rankify at ACL 2026 Demo.
Apr 05, 2026 ๐Ÿ—“๏ธ April 2026 โ€“ Three papers accepted at ACL 2026:
First paper (Main),
Second paper (Main), and
Third paper (Findings).
Apr 02, 2026 ๐Ÿ—“๏ธ April 2026 โ€“ Our Paper โ€œAre LLM-Based Retrievers Worth Their Cost?โ€ accepted at SIGIR 2026.
CORE A*

selected publications

OBLIQ-IR: Training a Dense Retriever for Oblique Queries

Abdalla M., Abdallah A., Sedek S., Jatowt A.
EMNLP 2026 (2026)

MM-BRIGHT: A Multi-Task Multimodal Benchmark for Reasoning-Intensive Retrieval

Abdallah A., Mounis M. D., Abdalla M., Kasem M. S., Senussi M. F., Mahmoud M., Ali M., Jatowt A., Kang H.-S.
KDD 2026 (2026)

Large Language Models Systematically Favor Popular Options: Evidence and Mitigation Across MCQs

Abdallah A., Ali M., Piryani B., Abdalla M., Jatowt A.
EMNLP 2026 (2026)