Set Learning for Generative Information Extraction

Jiangnan Li, Yice Zhang, Bin Liang, Kam-Fai Wong, Ruifeng Xu

Submission Type: Regular Short Paper

Submission Track: Information Extraction

Keywords: Information Extraction

TL;DR: We propose a task-agnostic approach called "Set Learning" to enhance generative information extraction by taking into account the unordered nature of structured tuples.

Abstract:

Recent efforts have endeavored to employ the sequence-to-sequence (Seq2Seq) model in Information Extraction~(IE) due to its potential to tackle multiple IE tasks in a unified manner.
Under this formalization, multiple structured objects are concatenated as the target sequence in a predefined order.
However, structured objects, by their nature, constitute an unordered set. Consequently, this formalization introduces a potential order bias, which can impair model learning.
Targeting this issue, this paper proposes a set learning approach that considers multiple permutations of structured objects to optimize set probability approximately. Notably, our approach does not require any modifications to model structures, making it easily integrated into existing generative IE frameworks. Experiments show that our method consistently improves existing frameworks on vast tasks and datasets.

Submission Number: 3681