Reproduce “Citation Recommendation based on Argumentative Zoning of User Queries” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shutian Ma, Chengzhi Zhang, Heng Zhang, Zheng Gao
A multi-task learning model was developed to enhance citation recommendation by integrating argumentative zoning of user queries. This model nearly doubled the recall for identifying true citations and achieved a higher F1-score compared to a single-task baseline, leveraging rhetorical intent from citing sentences.
Key insights
- The rhetorical structure and argumentative purpose embedded within citing sentences offer crucial semantic signals that can significantly improve citation recommendation.
- Multi-task learning successfully integrates argumentative zoning as an auxiliary task, leading to better generalization and performance in identifying relevant citations.
- Developing a specific argumentative zoning schema, such as the proposed Method/Conclusion/Goal/Object categories, is vital for effectively capturing citing intents.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.