Homa at SemEval-2025 Task 5: Aligning Librarian Records with OntoAligner for Subject Tagging

2025-04-30

AI Summary

This work addresses the automatic subject indexing of German technical literature records from the TIBKAT system, aligning them with the German authority classification scheme Gemeinsame Normdatei (GND). We propose a novel cross-lingual ontology alignment paradigm: for the first time, we adapt the OntoAligner framework to subject indexing by formalizing label assignment as a semantic alignment task between GND concepts and document descriptions. Our approach integrates multilingual semantic embeddings, retrieval-augmented generation (RAG)-enhanced candidate retrieval, and fine-grained similarity matching. Evaluated on SemEval-2025 Task 5, our method achieves significant improvements in GND category matching accuracy. It demonstrates strong robustness on German–English mixed records and high cross-lingual transferability. The framework provides a scalable, language-agnostic solution for automated knowledge organization of multilingual scientific literature.

Technology Category

• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Knowledge Representation and Reasoning: Ontologies
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

This paper presents our system, Homa, for SemEval-2025 Task 5: Subject Tagging, which focuses on automatically assigning subject labels to technical records from TIBKAT using the Gemeinsame Normdatei (GND) taxonomy. We leverage OntoAligner, a modular ontology alignment toolkit, to address this task by integrating retrieval-augmented generation (RAG) techniques. Our approach formulates the subject tagging problem as an alignment task, where records are matched to GND categories based on semantic similarity. We evaluate OntoAligner's adaptability for subject indexing and analyze its effectiveness in handling multilingual records. Experimental results demonstrate the strengths and limitations of this method, highlighting the potential of alignment techniques for improving subject tagging in digital libraries.

Problem

• Automatically assigning subject labels to technical records
• Aligning records to GND taxonomy using semantic similarity
• Evaluating OntoAligner for multilingual subject indexing

Innovation

• OntoAligner toolkit for ontology alignment
• Retrieval-augmented generation (RAG) techniques
• Semantic similarity for GND category matching