A Similarity Score Model for Aspect Category Detection

Zohreh Madhoushi,Abdul Razak Hamdan,Suhaila Zainudin

doi:10.14569/ijacsa.2021.0120738

Zohreh Madhoushi, Abdul Razak Hamdan + Show 1 more

Open Access

https://doi.org/10.14569/ijacsa.2021.0120738

Copy DOI

Abstract

Aspect-based Sentiment Analysis (ABSA) aims to extract significant aspects of an item or product from reviews and predict the sentiment of each aspect. Previous similarity methods tend to extract aspect categories at the word level by combining Language Models (LM) in their models. A drawback for the LM model is its dependence on a large amount of labelled data for a specific domain to function well. This work proposes a mechanism to address labelled data dependency by a one-step approach experimenting to decide the best combinatory architectures of recurrent-based LM and the best semantic similarity measures for fostering a new aspect category detection model. The proposed model addresses drawbacks of previous aspect category detection models in an implicit manner. The datasets of this study, S1 and S2, are from standard SemEval online competition. The proposed model outperforms the previous baseline models in terms of the F1-score of aspect category detection. This study finds more relevant aspect categories by creating a more stable and robust model. The F1 score of our best model for aspect category detection is 79.03% in the restaurant domain for the S1 dataset. In dataset S2, the F1-score is 72.65% in the laptop domain and 75.11% in the restaurant domain.

Highlights

The first task in Aspect-based Sentiment Analysis (ABSA) is to detect aspects of an item or product from reviews and categories each aspect into a specific group
Considering that deep learning model that used in the literature for this task are supervised and domain-dependent, this study aims to propose a mechanism for aspect category detection using sentence similarity measurement and recurrent-based Language Models (LM) without using labeled data
Another LM is trained on top of the initial LM with Amazon product review dataset in fourteen areas at the sentence level, and the model is fine-tuned for the in-domain dataset which is on laptop, restaurant and hotel

Summary

Introduction

The first task in Aspect-based Sentiment Analysis (ABSA) is to detect aspects of an item or product from reviews and categories each aspect into a specific group. A drastic advancement happened in the text representation, and many LMs were developed, such as Word2Vec [8], deep LM [9], [10] These emerging LMs have not yet fully addressed aspect category detection, mainly because there is no study to design experiments assessing the effect of different recent advanced LMs on the specific task of aspect category detection. Instead of learning from scratch using random weights, the representation created with the proposed LM can be a better starting point for another LM It can be used for a specific task in a more related domain. They cannot identify aspect categories directly from a review text Most of these model does not group the extracted terms into predefined categories in the literature [16], [18]. For the aspect category detection model to be practical, one crucial step is to propose a model that works in fewer steps

Objectives

Methods

Findings

Conclusion