Latent Variable Sentiment Grammar
This work addresses sentiment analysis for natural language processing, offering an incremental improvement by explicitly modeling sentiment composition in neural models.
The paper tackled the problem of sentiment classification over constituent trees by introducing sentiment grammar models that explicitly capture sentiment composition through latent variables and Gaussian mixture vectors, achieving state-of-the-art results on the Stanford Sentiment Treebank benchmark with ELMo embeddings.
Neural models have been investigated for sentiment classification over constituent trees. They learn phrase composition automatically by encoding tree structures but do not explicitly model sentiment composition, which requires to encode sentiment class labels. To this end, we investigate two formalisms with deep sentiment representations that capture sentiment subtype expressions by latent variables and Gaussian mixture vectors, respectively. Experiments on Stanford Sentiment Treebank (SST) show the effectiveness of sentiment grammar over vanilla neural encoders. Using ELMo embeddings, our method gives the best results on this benchmark.