Eliminating Sentiment Bias for Aspect-Level Sentiment Classification with Unsupervised Opinion Extraction

TL;DR

This paper introduces an unsupervised opinion extraction method that eliminates sentiment bias in aspect representations for improved aspect-level sentiment classification, achieving state-of-the-art results.

Contribution

It proposes a span-based anti-bias framework that removes sentiment bias and aligns opinion candidates, enhancing interpretability and performance in ALSC.

Findings

Achieves state-of-the-art results on five benchmarks.

Effectively extracts opinions without supervision.

Reduces sentiment bias in aspect representations.

Abstract

Aspect-level sentiment classification (ALSC) aims at identifying the sentiment polarity of a specified aspect in a sentence. ALSC is a practical setting in aspect-based sentiment analysis due to no opinion term labeling needed, but it fails to interpret why a sentiment polarity is derived for the aspect. To address this problem, recent works fine-tune pre-trained Transformer encoders for ALSC to extract an aspect-centric dependency tree that can locate the opinion words. However, the induced opinion words only provide an intuitive cue far below human-level interpretability. Besides, the pre-trained encoder tends to internalize an aspect's intrinsic sentiment, causing sentiment bias and thus affecting model performance. In this paper, we propose a span-based anti-bias aspect representation learning framework. It first eliminates the sentiment bias in the aspect embedding by adversarial learning against aspects' prior sentiment. Then, it aligns the distilled opinion candidates with the aspect by span-based dependency modeling to highlight the interpretable opinion terms. Our method achieves new state-of-the-art performance on five benchmarks, with the capability of unsupervised opinion extraction.