Learning Neural Textual Representations for Citation Recommendation
This addresses the challenge of manually selecting citations for scientific papers, offering an incremental improvement in automated recommendation systems for researchers and authors.
The paper tackles the problem of automated citation recommendation by proposing a novel approach that combines deep sequential document representations (Sentence-BERT) with Siamese and triplet networks in a submodular scoring function, achieving state-of-the-art results on the ACL Anthology Network corpus by outperforming all baselines in metrics like MRR and F1-at-k.
With the rapid growth of the scientific literature, manually selecting appropriate citations for a paper is becoming increasingly challenging and time-consuming. While several approaches for automated citation recommendation have been proposed in the recent years, effective document representations for citation recommendation are still elusive to a large extent. For this reason, in this paper we propose a novel approach to citation recommendation which leverages a deep sequential representation of the documents (Sentence-BERT) cascaded with Siamese and triplet networks in a submodular scoring function. To the best of our knowledge, this is the first approach to combine deep representations and submodular selection for a task of citation recommendation. Experiments have been carried out using a popular benchmark dataset - the ACL Anthology Network corpus - and evaluated against baselines and a state-of-the-art approach using metrics such as the MRR and F1-at-k score. The results show that the proposed approach has been able to outperform all the compared approaches in every measured metric.