BERT-ERC: Fine-Tuning BERT Is Enough for Emotion Recognition in Conversation

Authors: Xiangyu Qin, Zhiyu Wu, Tingting Zhang, Yanran Li, Jian Luan, Bin Wang, Li Wang, Jinshi Cui

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Compared to existing methods, BERT-ERC achieves substantial improvement on four datasets, indicating its effectiveness and generalization capability. Besides, we also set up the limited resources scenario and the online prediction scenario to approximate real-world scenarios. Extensive experiments demonstrate that the proposed paradigm significantly outperforms the previous one and can be adapted to various scenes.
Researcher Affiliation Collaboration 1School of Intelligence Science and Technology, Peking University 2Xiaomi AI Lab 3School of Psychological and Cognitive Sciences and Beijing Key Laboratory of Behavior and Mental Health, Peking University
Pseudocode No The paper describes its methods in text and figures but does not include a clearly labeled 'Pseudocode' or 'Algorithm' block or section.
Open Source Code No The paper does not provide any explicit statement about releasing source code or a link to a code repository for the described methodology.
Open Datasets Yes Our experiments involve four datasets, whose information is as follows. IEMOCAP (Busso et al. 2008) is a multi-modal dataset... MELD (Poria et al. 2018) is a multi-modal dataset extracted from the TV show Friends. Daily Dialog (Li et al. 2017) collects conversations of English learners. Emory NLP (Zahiri and Choi 2018) is also built on the TV show Friends...
Dataset Splits Yes Following previous works, dialogues of the first four sessions are used as the training set and the rest are used as the test set. (for IEMOCAP)
Hardware Specification Yes We implement all experiments on 4 NVIDIA Tesla V100 GPUs with the Pytorch framework.
Software Dependencies No The paper mentions using the 'Pytorch framework' but does not specify a version number for it or any other key software dependencies required for reproduction.
Experiment Setup Yes We use the Adam optimizer (Kingma and Ba 2014) with a learning rate of 9e-6 in experiments. Besides, we utilize a 1-layer MLP as the classifier unless otherwise specified. For all datasets, we train 10 epochs with the batch size of 8. Focal Loss (Lin et al. 2017) is applied to alleviate the class imbalance problem.