Team LA at SCIDOCA shared task 2025: Citation Discovery via relation-based zero-shot retrieval
This work addresses citation discovery for academic document analysis, but it is incremental as it builds on existing retrieval and LLM methods.
The paper tackled the problem of predicting the correct citation from a candidate pool for a given paragraph, addressing challenges like paragraph length and high similarity among candidates, and achieved effectiveness in citation prediction as evaluated on the SCIDOCA 2025 training dataset.
The Citation Discovery Shared Task focuses on predicting the correct citation from a given candidate pool for a given paragraph. The main challenges stem from the length of the abstract paragraphs and the high similarity among candidate abstracts, making it difficult to determine the exact paper to cite. To address this, we develop a system that first retrieves the top-k most similar abstracts based on extracted relational features from the given paragraph. From this subset, we leverage a Large Language Model (LLM) to accurately identify the most relevant citation. We evaluate our framework on the training dataset provided by the SCIDOCA 2025 organizers, demonstrating its effectiveness in citation prediction.