Abstract

The e-commerce has started a new trend in natural language processing throughsentiment analysis of user-generated reviews. Different consumers havedifferent concerns about various aspects of a specific product or service.Aspect category detection, as a subtask of aspect-based sentiment analysis,tackles the problem of categorizing a given review sentence into a set ofpre-defined aspect categories. In recent years, deep learning approaches havebrought revolutionary advances in multiple branches of natural languageprocessing including sentiment analysis. In this paper, we propose a deepneural network method based on attention mechanism to identify different aspectcategories of a given review sentence. Our model utilizes several attentionswith different topic contexts, enabling it to attend to different parts of areview sentence based on different topics. Experimental results on two datasetsin the restaurant domain released by SemEval workshop demonstrates that ourapproach outperforms existing methods on both datasets. Visualization of thetopic attention weights shows the effectiveness of our model in identifyingwords related to different topics.

 

Introduction (beta)

None

 

Conclusion (beta)

None