ISCAS at SemEval-2020 Task 5: Pre-trained Transformers for Counterfactual Statement Modeling

TL;DR

This paper presents a transformer-based system for detecting counterfactual statements and extracting antecedents and consequences, achieving third place in both SemEval-2020 Task 5 subtasks.

Contribution

It introduces a novel application of pre-trained transformers to counterfactual statement modeling and formulates antecedent and consequence extraction as a question answering task.

Findings

Both subtasks achieved third place in evaluation.

Transformer classifiers effectively detect counterfactual statements.

Question answering approach successfully extracts antecedent and consequence.

Abstract

ISCAS participated in two subtasks of SemEval 2020 Task 5: detecting counterfactual statements and detecting antecedent and consequence. This paper describes our system which is based on pre-trained transformers. For the first subtask, we train several transformer-based classifiers for detecting counterfactual statements. For the second subtask, we formulate antecedent and consequence extraction as a query-based question answering problem. The two subsystems both achieved third place in the evaluation. Our system is openly released at https://github.com/casnlu/ISCAS-SemEval2020Task5.