SOUL: Towards Sentiment and Opinion Understanding of Language
This work addresses the problem of improving sentiment analysis for natural language processing researchers by introducing a more challenging task and dataset, though it is incremental as it builds on existing sentiment analysis frameworks.
The authors tackled the limitation of existing sentiment analysis models in capturing broader complexities by proposing a new task called Sentiment and Opinion Understanding of Language (SOUL), which includes subtasks for review comprehension and justification generation, and they created a dataset of 15,028 statements, finding a performance gap of up to 27% between models and humans.
Sentiment analysis is a well-established natural language processing task, with sentiment polarity classification being one of its most popular and representative tasks. However, despite the success of pre-trained language models in this area, they often fall short of capturing the broader complexities of sentiment analysis. To address this issue, we propose a new task called Sentiment and Opinion Understanding of Language (SOUL). SOUL aims to evaluate sentiment understanding through two subtasks: Review Comprehension (RC) and Justification Generation (JG). RC seeks to validate statements that focus on subjective information based on a review text, while JG requires models to provide explanations for their sentiment predictions. To enable comprehensive evaluation, we annotate a new dataset comprising 15,028 statements from 3,638 reviews. Experimental results indicate that SOUL is a challenging task for both small and large language models, with a performance gap of up to 27% when compared to human performance. Furthermore, evaluations conducted with both human experts and GPT-4 highlight the limitations of the small language model in generating reasoning-based justifications. These findings underscore the challenging nature of the SOUL task for existing models, emphasizing the need for further advancements in sentiment analysis to address its complexities. The new dataset and code are available at https://github.com/DAMO-NLP-SG/SOUL.