Combination of Convolutional Neural Network and Gated Recurrent Unit for Aspect-Based Sentiment Analysis

Aspect-based sentiment analysis (ABSA) aims to identify views and polarities towards a given aspect in reviews. Compared with general analysis, ABSA can provide more detailed complete information. Recently, has become an important task for natural language understanding attracted considerable attention from both academic industry fields. The polarity of sentence is not only decided by its content but also relatively significant correlation the targeted aspect. For this reason, we propose model aspect-based which combination Convolutional Neural Network (CNN) Gated Recurrent Unit (GRU), utilizing local features generated CNN long-term dependency learned GRU. Extensive experiments have been conducted on datasets hotels cars, results show that proposed achieves excellent performance terms extraction classification. Experiments demonstrate great domain expansion capability model.