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

Chi Sun, Luyao Huang, Xipeng Qiu
3/22/2019
cs.CL

Official code: HSLCY/ABSA-BERT-pair (519 stars, MIT, last commit 2022-01-04).

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.

View on arXivView PDF

Code Implementations(1)

HSLCY/ABSA-BERT-pairOfficial100%
MIT

Utilizing BERT for Aspect-Based Sentiment Analysis via Constructing Auxiliary Sentence (NAACL 2019)

519133Shell, PythonMar 22, 20194 years agoMIT

Similar Papers

Extracting Emotion Phrases from Tweets using BART

87% match2024

Deep Learning Brasil at ABSAPT 2022: Portuguese Transformer Ensemble Approaches

86% match1 repo42023

USA: Universal Sentiment Analysis Model & Construction of Japanese Sentiment Text Classification and Part of Speech Dataset

86% match2023

Multiple Instance Learning Networks for Fine-Grained Sentiment Analysis

86% match1 repo352017

SOUL: Towards Sentiment and Opinion Understanding of Language

86% match1 repo72023

A Multi-task Learning Model for Chinese-oriented Aspect Polarity Classification and Aspect Term Extraction

85% match1 repo1982019

Cite this paper

@article{sun2019utilizing,
  title  = {Utilizing BERT for Aspect-Based Sentiment Analysis via Constructing Auxiliary Sentence},
  author = {Chi Sun and Luyao Huang and Xipeng Qiu},
  year   = {2019},
  eprint = {1903.09588},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1903.09588}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1903.09588.svg)](https://codeofpaper.com/paper/1903.09588)

Discussion