BERT (Bidirectional Encoder Representations from Transformers) uses an encoder architecture with an attention mechanism to construct a transformer-based neural network. In this study, we develop a Chinese word-level BERT to learn contextual language representations and propose a transformer fusion framework for Chinese sentiment intensity prediction in the valence-arousal dimensions. Experimental results on the Chinese EmoBank indicate that our transformer-based fusion model outperforms other neuralnetwork-based, regression-based and lexicon-based methods, reflecting the effectiveness of integrating semantic representations in different degrees of linguistic granularity. Our proposed transformer fusion framework is also simple and easy to fine-tune over different downstream tasks.
INDEX TERMSTransformer fusion, Chinese word-level BERT, pre-trained language models, dimensional sentiment analysis, affective computing.