Semantically Orthogonal Framework for Citation Classification: Disentangling Intent and Content

📅 2026-01-08
🏛️ International Conference on Theory and Practice of Digital Libraries
📈 Citations: 1
Influential: 0
📄 PDF
🤖 AI Summary
This study addresses the conflation of citation intent and cited content type in existing citation classification methods, which hinders both fine-grained distinction and classification reliability. To resolve this, the authors propose SOFT, a novel framework grounded in semantic role theory that decouples citation intent and content type into an orthogonal two-dimensional classification scheme, establishing clear and reusable annotation guidelines. Through manual re-annotation of the ACL-ARC dataset and the creation of a cross-disciplinary ACT2 test set, the authors evaluate SOFT using both zero-shot and fine-tuned large language models. Results demonstrate that SOFT significantly outperforms existing frameworks such as ACL-ARC and SciCite in enhancing agreement between human annotators and models, improving classification performance, and enabling robust cross-domain generalization.

Technology Category

Application Category

📝 Abstract
Understanding the role of citations is essential for research assessment and citation-aware digital libraries. However, existing citation classification frameworks often conflate citation intent (why a work is cited) with cited content type (what part is cited), limiting their effectiveness in auto classification due to a dilemma between fine-grained type distinctions and practical classification reliability. We introduce SOFT, a Semantically Orthogonal Framework with Two dimensions that explicitly separates citation intent from cited content type, drawing inspiration from semantic role theory. We systematically re-annotate the ACL-ARC dataset using SOFT and release a cross-disciplinary test set sampled from ACT2. Evaluation with both zero-shot and fine-tuned Large Language Models demonstrates that SOFT enables higher agreement between human annotators and LLMs, and supports stronger classification performance and robust cross-domain generalization compared to ACL-ARC and SciCite annotation frameworks. These results confirm SOFT's value as a clear, reusable annotation standard, improving clarity, consistency, and generalizability for digital libraries and scholarly communication infrastructures. All code and data are publicly available on GitHub https://github.com/zhiyintan/SOFT.
Problem

Research questions and friction points this paper is trying to address.

citation classification
citation intent
cited content type
semantic disentanglement
annotation framework
Innovation

Methods, ideas, or system contributions that make the work stand out.

citation classification
semantic orthogonality
intent-content disentanglement
annotation framework
cross-domain generalization
💼 Related Jobs
No related jobs found.
C
Changxu Duan
Technische Universität Darmstadt, Darmstadt, Germany
Z
Zhiyin Tan
L3S Research Center, Leibniz University Hannover, Hannover, Germany