ARC-NLP at PAN 2023: Transition-Focused Natural Language Inference for Writing Style Detection

TL;DR

This paper presents a transition-focused natural language inference approach using Transformer models, specifically DeBERTa, to detect writing style changes between consecutive paragraphs, outperforming baselines in multi-author style detection.

Contribution

The paper introduces a novel transition-focused NLI formulation for writing style change detection and demonstrates the effectiveness of Transformer-based models with warmup training in this task.

Findings

DeBERTa with warmup training outperforms baselines.

Transition-focused NLI improves style change detection.

Model performs well across different difficulty setups.

Abstract

The task of multi-author writing style detection aims at finding any positions of writing style change in a given text document. We formulate the task as a natural language inference problem where two consecutive paragraphs are paired. Our approach focuses on transitions between paragraphs while truncating input tokens for the task. As backbone models, we employ different Transformer-based encoders with warmup phase during training. We submit the model version that outperforms baselines and other proposed model versions in our experiments. For the easy and medium setups, we submit transition-focused natural language inference based on DeBERTa with warmup training, and the same model without transition for the hard setup.