Multi-Aspect Sentiment Analysis with Latent Sentiment-Aspect Attribution

TL;DR

This paper presents SAAM, a novel framework that enhances multi-aspect sentiment analysis by leveraging sentence-embedding correlations, improving performance and providing deeper data insights.

Contribution

Introduces the sentiment-aspect attribution module (SAAM), a new framework that improves multi-aspect sentiment classification and offers interpretability over traditional neural models.

Findings

SAAM improves sentiment analysis accuracy on hotel and beer review datasets.

The framework enables semi-supervised sentence aspect labeling.

SAAM provides potential for sentiment snippet extraction.

Abstract

In this paper, we introduce a new framework called the sentiment-aspect attribution module (SAAM). SAAM works on top of traditional neural networks and is designed to address the problem of multi-aspect sentiment classification and sentiment regression. The framework works by exploiting the correlations between sentence-level embedding features and variations of document-level aspect rating scores. We demonstrate several variations of our framework on top of CNN and RNN based models. Experiments on a hotel review dataset and a beer review dataset have shown SAAM can improve sentiment analysis performance over corresponding base models. Moreover, because of the way our framework intuitively combines sentence-level scores into document-level scores, it is able to provide a deeper insight into data (e.g., semi-supervised sentence aspect labeling). Hence, we end the paper with a detailed analysis that shows the potential of our models for other applications such as sentiment snippet extraction.