One2Set: Generating Diverse Keyphrases as a Set

TL;DR

This paper introduces One2Set, a novel approach for keyphrase generation that treats keyphrases as an unordered set, using parallel generation with control codes and bipartite matching to improve diversity and accuracy.

Contribution

The paper proposes a new training paradigm and model that generate keyphrases as an unordered set, avoiding order bias and enhancing diversity compared to previous sequence-based methods.

Findings

Outperforms state-of-the-art methods on multiple benchmarks.

Increases diversity and reduces duplication in generated keyphrases.

Effectively models keyphrases as an unordered set.

Abstract

Recently, the sequence-to-sequence models have made remarkable progress on the task of keyphrase generation (KG) by concatenating multiple keyphrases in a predefined order as a target sequence during training. However, the keyphrases are inherently an unordered set rather than an ordered sequence. Imposing a predefined order will introduce wrong bias during training, which can highly penalize shifts in the order between keyphrases. In this work, we propose a new training paradigm One2Set without predefining an order to concatenate the keyphrases. To fit this paradigm, we propose a novel model that utilizes a fixed set of learned control codes as conditions to generate a set of keyphrases in parallel. To solve the problem that there is no correspondence between each prediction and target during training, we propose a $K$-step target assignment mechanism via bipartite matching, which greatly increases the diversity and reduces the duplication ratio of generated keyphrases. The experimental results on multiple benchmarks demonstrate that our approach significantly outperforms the state-of-the-art methods.