1Cademy @ Causal News Corpus 2022: Enhance Causal Span Detection via Beam-Search-based Position Selector

TL;DR

This paper introduces a novel span detection method for causal signal extraction using a reading comprehension framework, data augmentation, and beam-search post-processing, achieving top performance in the CASE 2022 competition.

Contribution

The paper presents a new approach combining RC-based span prediction, data augmentation, and beam-search strategy for improved causal span detection.

Findings

Achieved 54.15 F1 score, ranking 1st in CASE 2022.

Demonstrated effectiveness of data augmentation and beam-search in span detection.

Proposed a scalable post-processing method for span detection tasks.

Abstract

In this paper, we present our approach and empirical observations for Cause-Effect Signal Span Detection -- Subtask 2 of Shared task 3~\cite{tan-etal-2022-event} at CASE 2022. The shared task aims to extract the cause, effect, and signal spans from a given causal sentence. We model the task as a reading comprehension (RC) problem and apply a token-level RC-based span prediction paradigm to the task as the baseline. We explore different training objectives to fine-tune the model, as well as data augmentation (DA) tricks based on the language model (LM) for performance improvement. Additionally, we propose an efficient beam-search post-processing strategy to due with the drawbacks of span detection to obtain a further performance gain. Our approach achieves an average $F_1$ score of 54.15 and ranks \textbf{$1^{st}$} in the CASE competition. Our code is available at \url{https://github.com/Gzhang-umich/1CademyTeamOfCASE}.