Discriminative Models Still Outperform Generative Models in Aspect Based Sentiment Analysis In Cross-Domain and Cross-Lingual Settings

Anonymous

Abstract: Aspect-based Sentiment Analysis (ABSA) helps to explain customers' opinions towards products and services. In the past, ABSA models were discriminative, but more recently generative models have been used to generate aspects and polarities directly from text. In contrast, discriminative models first select aspects from the text, and then classify the aspect's polarity. Previous results showed that generative models outperform discriminative models on several English ABSA datasets. Here, we rigorously contrast discriminative and generative models in several settings. We compare both model types in cross-lingual, cross-domain, and cross- lingual and domain, to understand generalizability in settings other than mono-lingual English in-domain. Our more thorough evaluation shows that, contrary to previous studies, discriminative models still clearly outperform generative models in almost all settings.