A Comparative Study of Different Pre-trained Language Models for Sentiment Analysis of Human-Computer Negotiation Dialogue

This paper offers a comprehensive comparative study of various pre-trained language models for sentiment analysis in human-computer negotiation dialogues. It examines numerous state-of-the-art PLMs, including GPT-3.5, BERT, and its variants, along with other models like Claude, ELECTRA, NEZHA, ERNIE 3.0, BART, and XLNet, focusing particularly on their effectiveness in sentiment detection in negotiation dialogues. Using a large, diverse dataset annotated with sentiment labels, the study assesses these models using accuracy, precision, recall, and F1 metrics. The findings highlight distinct performance differences among the models, providing insights for future research in automated negotiation systems and sentiment analysis in this context.