Publications

A collection of my research publications and contributions to the field of Computer Science.

On the impact of Compression and Aggregation on Knowledge Graph Embeddings

Authors: Nader Sobhi, Unmesh Joshi, Jacopo Urbani, Katja Hose
Venue: KG-NeSy'26 @ International Semantic Web Conference 2026

In this paper we evaluate how two approaches that attempt to optimize different issues with the use of KGEs effect link prediction performance. We then empirically study the amount by which the embeddings can be modified before severe degradation in link prediction performance. Based on our results we give method agnostic thresholds for how much embeddings can deviate before this degradation occurs.

DORSET: Decoding hOw Knowledge Graph chaRacteristics Shape Embedding sTrategies

Authors: Majlinda Llugiqi, Fajar J. Ekaputra, Marta Sabou, Nader Sobhi, Hesham Morgan, Eleonora Laurenza, Milos Jovanovik, Katja Hose, Emanuel Sallinger
Venue: KG-NeSy'26 @ International Semantic Web Conference 2026

The DORSET project (https://dorset-project.org/) aims to address this gap from two aspects: (1) building a systematic, experimentally grounded understanding of how KG characteristics shape KGE performance, and (2) using these insights to advance both KGE and knowledge engineering (KE) research. Expected outcomes include diverse benchmarks, KG-characteristic-aware embedding methods (KGE), and new KG engineering methodologies and tools (KE).