Fine-Tuning BERT for Sentiment Analysis of Vietnamese Reviews
This work addresses the problem of sentiment analysis for Vietnamese reviews, providing incremental improvements for researchers and practitioners working with this specific language.
This paper fine-tunes BERT for sentiment analysis of Vietnamese reviews, comparing two methods: one using only the [CLS] token and another using all BERT output vectors. The BERT-based models slightly outperform models using GloVe and FastText, and the proposed fine-tuning method shows better performance than the original BERT fine-tuning method on the datasets used.
Sentiment analysis is an important task in the field ofNature Language Processing (NLP), in which users' feedbackdata on a specific issue are evaluated and analyzed. Manydeep learning models have been proposed to tackle this task, including the recently-introduced Bidirectional Encoder Rep-resentations from Transformers (BERT) model. In this paper,we experiment with two BERT fine-tuning methods for thesentiment analysis task on datasets of Vietnamese reviews: 1) a method that uses only the [CLS] token as the input for anattached feed-forward neural network, and 2) another methodin which all BERT output vectors are used as the input forclassification. Experimental results on two datasets show thatmodels using BERT slightly outperform other models usingGloVe and FastText. Also, regarding the datasets employed inthis study, our proposed BERT fine-tuning method produces amodel with better performance than the original BERT fine-tuning method.