Utilizing BERT for Aspect-Based Sentiment Analysis via Constructing Auxiliary Sentence

TL;DR

This paper improves aspect-based sentiment analysis by transforming it into a sentence-pair classification task using auxiliary sentences and fine-tuning BERT, achieving state-of-the-art results.

Contribution

The novel approach constructs auxiliary sentences from aspects to reformulate ABSA as a sentence-pair task, enhancing BERT's effectiveness.

Findings

Achieved new state-of-the-art results on SentiHood dataset.

Achieved new state-of-the-art results on SemEval-2014 Task 4.

Demonstrated the effectiveness of auxiliary sentence construction for ABSA.

Abstract

Aspect-based sentiment analysis (ABSA), which aims to identify fine-grained opinion polarity towards a specific aspect, is a challenging subtask of sentiment analysis (SA). In this paper, we construct an auxiliary sentence from the aspect and convert ABSA to a sentence-pair classification task, such as question answering (QA) and natural language inference (NLI). We fine-tune the pre-trained model from BERT and achieve new state-of-the-art results on SentiHood and SemEval-2014 Task 4 datasets.