Polarity and Intensity: the Two Aspects of Sentiment Analysis
Leimin Tian, Catherine Lai, Johanna D. Moore · 2018
Current multimodal sentiment analysis frames sentiment score prediction as a general Machine Learning task.However, what the sentiment score actually represents has often been overlooked.As a measurement of opinions and affective states, a sentiment score generally consists of two aspects: polarity and intensity.We decompose sentiment scores into these two aspects and study how they are conveyed through individual modalities and combined multimodal models in a naturalistic monologue setting.In particular, we build unimodal and multimodal multitask learning models with sentiment score prediction as the main task and polarity and/or intensity classification as the auxiliary tasks.Our experiments show that sentiment analysis benefits from multi-task learning, and individual modalities differ when conveying the polarity and intensity aspects of sentiment.