ERU-KG: Efficient Reference-aligned Unsupervised Keyphrase Generation

ACL ARR 2025 February Submission1241 Authors

Abstract: Unsupervised keyphrase prediction has gained growing interest in recent years. However, existing methods typically rely on heuristically defined importance scores, which may lead to inaccurate informativeness estimation. In addition, they lack consideration for time efficiency. To solve these problems, we propose ERU-KG, an unsupervised keyphrase generation (UKG) model that consists of a phraseness and an informativeness module. The former generate candidates, while the latter estimate their relevance. The informativeness module innovates by learning to model informativeness through references (e.g., queries, citation contexts, and titles) and at the term-level, thereby 1) capturing how the key concepts of the document are perceived in different contexts and 2) estimate informativeness of phrases more efficiently by aggregating term informativeness, removing the need for explicit modeling of the candidates. ERU-KG demonstrates its effectiveness on keyphrase generation benchmarks by outperforming unsupervised baselines and achieving on average 89% of the performance of a supervised baseline for top 10 predictions. Additionally, to highlight its practical utility, we evaluate the model on text retrieval tasks and show that keyphrases generated by ERU-KG are effective when employed as query and document expansions. Finally, inference speed tests reveal that ERU-KG is the fastest among baselines of similar model sizes.

Paper Type: Long

Research Area: Information Extraction

Research Area Keywords: document-level extraction, zero/few-shot extraction

Contribution Types: Approaches to low-resource settings, Approaches low compute settings-efficiency

Languages Studied: English

Submission Number: 1241