A Deep Learning Approach to Classify Aspect-Level Sentiment using Small Datasets
João Paulo Aires, Carlos Eduardo de Araújo Padilha, Christian Vahl Quevedo, Felipe Rech Meneguzzi · 2018
Sentiment analysis is an important technique to interpret user opinion on products from text, for example, as shared in social media. Recent approaches using deep learning can accurately extract overall sentiment from large datasets. However, extracting sentiment from specific aspects of a product with small training datasets remains a challenge. The automatic classification of sentiments at aspect-level can provide more detailed feedbacks about product and service opinions avoiding manual verification. In this work, we develop two deep learning approaches to classify sentiment at aspect-level using small datasets.