Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Information Retrieval
Title: Local Citation Recommendation with Hierarchical-Attention Text Encoder and SciBERT-based Reranking
(Submitted on 2 Dec 2021 (v1), last revised 17 Mar 2022 (this version, v3))
Abstract: The goal of local citation recommendation is to recommend a missing reference from the local citation context and optionally also from the global context. To balance the tradeoff between speed and accuracy of citation recommendation in the context of a large-scale paper database, a viable approach is to first prefetch a limited number of relevant documents using efficient ranking methods and then to perform a fine-grained reranking using more sophisticated models. In that vein, BM25 has been found to be a tough-to-beat approach to prefetching, which is why recent work has focused mainly on the reranking step. Even so, we explore prefetching with nearest neighbor search among text embeddings constructed by a hierarchical attention network. When coupled with a SciBERT reranker fine-tuned on local citation recommendation tasks, our hierarchical Attention encoder (HAtten) achieves high prefetch recall for a given number of candidates to be reranked. Consequently, our reranker requires fewer prefetch candidates to rerank, yet still achieves state-of-the-art performance on various local citation recommendation datasets such as ACL-200, FullTextPeerRead, RefSeer, and arXiv.
Submission history
From: Nianlong Gu [view email][v1] Thu, 2 Dec 2021 13:20:26 GMT (435kb,D)
[v2] Thu, 3 Mar 2022 10:19:33 GMT (225kb,D)
[v3] Thu, 17 Mar 2022 08:31:22 GMT (225kb,D)
Link back to: arXiv, form interface, contact.