Learning affective meanings that derives the social behavior using Bidirectional Encoder Representations from Transformers
This work addresses the prohibitive cost of survey data in social science research by providing a scalable alternative, though it is incremental as it applies an existing method to a new domain.
The paper tackles the problem of quantifying sentiments for social behavior prediction by replacing costly surveys with a fine-tuned BERT model, achieving state-of-the-art accuracy in estimating affective meanings and expanding the lexicon.
Predicting the outcome of a process requires modeling the system dynamic and observing the states. In the context of social behaviors, sentiments characterize the states of the system. Affect Control Theory (ACT) uses sentiments to manifest potential interaction. ACT is a generative theory of culture and behavior based on a three-dimensional sentiment lexicon. Traditionally, the sentiments are quantified using survey data which is fed into a regression model to explain social behavior. The lexicons used in the survey are limited due to prohibitive cost. This paper uses a fine-tuned Bidirectional Encoder Representations from Transformers (BERT) model to develop a replacement for these surveys. This model achieves state-of-the-art accuracy in estimating affective meanings, expanding the affective lexicon, and allowing more behaviors to be explained.