Embarrassingly Simple Unsupervised Aspect Extraction
This provides a straightforward and effective solution for aspect extraction in sentiment analysis, applicable across domains and languages, though it is incremental in simplifying existing approaches.
The authors tackled aspect identification in sentiment analysis by introducing Contrastive Attention (CAt), a simple unsupervised method using word embeddings and a POS tagger, which achieved a considerable performance boost and interpretability.
We present a simple but effective method for aspect identification in sentiment analysis. Our unsupervised method only requires word embeddings and a POS tagger, and is therefore straightforward to apply to new domains and languages. We introduce Contrastive Attention (CAt), a novel single-head attention mechanism based on an RBF kernel, which gives a considerable boost in performance and makes the model interpretable. Previous work relied on syntactic features and complex neural models. We show that given the simplicity of current benchmark datasets for aspect extraction, such complex models are not needed. The code to reproduce the experiments reported in this paper is available at https://github.com/clips/cat