Emotion recognition in conversations (ERC) aims to predict emotional categories of utterances within a conversation. Presently, graph neural network-based ERC methods predominantly employ fixed hyperparameters to determine the connections among graph edges, lacking adaptive strategies for edge construction tailored to diverse data and ignoring thematic relationships between statements. Furthermore, during the training process of graph neural networks, these methods often utilize a summation superposition approach to aggregate node information, limiting the model's non-linear capabilities. To address these limitations, this paper integrated topic modeling with graph neural networks and proposed a novel edge construction method. Firstly, a topic model was employed to extract the thematic distribution of statements within a conversation, followed by the connection of statements sharing same themes. Meanwhile, the SwiGLU gated unit was introduced to regulate the flow of information between layers in the graph neural network. Considering differences in character information, the edge types were carefully tailored to better capture intrinsic and extrinsic factors influencing emotional changes. Through extensive experiments conducted on four publicly available datasets (IEMOCAP, MELD, EmoryNLP, DailyDialogue), our approach demonstrates significant improvements over advanced ERC methods, achieving F1 score enhancements of 1.69%, 0.27%, and 0.38% on the first three datasets, respectively. Moreover, our adaptive method exhibits a 2.11% improvement on long conversations, surpassing the 0.8% gain on short conversations. The introduction of the SwiGLU unit effectively mitigates over-smoothing phenomena in the graph neural network. Consequently, the proposed approach, which combines adaptive graph construction with topic modeling and integrates SwiGLU gated units into graph neural networks, proves to be effective in enhancing dialogue emotion recognition, thereby reinforcing the model’s robustness.
HONGWei, LIMin. A review: Text sentiment analysis methods[J].Computer Engineering and Science, 2019, 41(4): 750-757. (in Chinese)
[3]
李思贤. 基于恐惧情感强度计算的自动驾驶车辆决策机制研究[D].青岛: 山东科技大学, 2020.
[4]
MAJUMDERN, HONGP, PENGS, et al. MIME: MIMicking emotions for empathetic response generation[C]//Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020: 8968-8979.
[5]
DAUPHINY N, FANA, AULIM, et al.Language modeling with gated convolutional networks[C]//International Conference on Machine Learning, PMLR, 2017: 933-941.
[6]
LANS, MAY, HUANGW, et al. Dstagnn: Dynamic spatial-temporal aware graph neural network for traffic flow forecasting[C]//International Conference on Machine Learning, PMLR, 2022: 11906-11917.
HAZARIKAD, PORIAS, MIHALCEAR, et al. Icon: Inter- active conversational memory network for multimodal emotion detection[C]//Conference on Empirical Methods in Natural Language Processing, 2018: 2594-2604.
[9]
HAZARIKAD, PORIAS, ZADEHA, et al.Conversational memory network for emotion recognition in dyadic dialogue videos[C]//Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. New Orleans, 2018: 2122-2132.
[10]
MAJUMDERN, PORIAS, HAZARIKAD, et al.Dialogue-RNN: An attentive rnn for emotion detection in conversations[C]//AAAI Conference on Artificial Intelligence, 2019: 6818-6825.
[11]
ZHUANGC, MAQ. Dual graph convolutional networks for graph-based semi-supervised classification[C]//World Wide Web Conference, 2018: 499-508.
[12]
GHOSALD, MAJUMDERN, PORIAS, et al.Dialogue-GCN: A graph convolutional neural network for emotion recognition in conversation[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP),2020: 154‑164.
[13]
DEVLINJ, CHANGM W, LEEK, et al. BERT: Pre-training of deep bidirectional transformers for language understanding[C]//Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2019: 4171-4186.
[14]
LIUY H, OTT M, GOYALN, et al. RoBERTa: A robustly optimized BERT pretraining approach[DB/OL].(2019-07-26)[2023-11-15].
[15]
ISHIWATARIT, YASUDAY, MIYAZAKIT, et al.Relati-on aware graph attention networks with relational position encodings for emotion recognition in conversations[C]//Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020: 7360-7370.
[16]
ZHANGD, WUL Q, SUNC L, et al. Modeling both context-and speaker-sensitive dependence for emotion detection in multi-speaker conversation[C]//28th International Joint Conference on Artificial Intelligence, 2019: 5415-5421.
[17]
SHENW, WUS, YANGY, et al. Directed acyclic graph network for conversational emotion recogniton[C]//The 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, 2021: 1551-1560.
[18]
JELODARH, WANGY, YUANC, et al. Latent Dirichlet allocation (LDA) and topic modeling: models, applications, a survey[J]. Multimedia Tools and Applications, 2019, 78(11): 15169-15211.
[19]
VELIČKOVIĆP, CUCURULLG, CASANOVAA, et al. Graph attention networks[DB/OL].(2017-10-30)[2023-11-15].
[20]
YANGC Q, WANGR J, YAOS C, et al. Revisiting over-smoothing in deep GCNs[DB/OL]. (2020-03-30)[2023-11-15].
[21]
BUSSOC, BULUTM, LEEC C, et al. IEMOCAP: Interactive emotional dyadic motion capture database[J]. Language Resources and Evaluation, 2008, 42: 335-359.
[22]
PORIAS, HAZARIKAD, MAJUMDERN, et al. MELD: A Multimodal multi-party dataset for emotion recognition in conversations[DB/OL]. (2018-10-05)[2023-11-15].
[23]
LIY R, SUH, SHENX Y, et al. DailyDialog: A manu-ally labelled multiturn dialogue dataset[DB/OL].(2017-10-11)[2023-11-15].
[24]
ZAHIRIS M, CHOIJ D. Emotion detection on TV show transcripts with sequence-based convolutional neural networks[DB/OL].(2017-08-14)[2023-11-15].
[25]
GHOSALD, MAJUMDERN, GELBUKHA, et al. COSMIC: Common sense knowledge for emotion identification in conversations[C]//Findings of the Association for Computational Linguistics Findings of ACL: EMNLP 2020. Association for Computational Linguistics (ACL), 2020: 2470-2481.
[26]
SUNY, YUN, FUG. A discourse-aware graph neural network for emotion recognition in multi-party conversation[C]//Findings of the Association for Computational Linguistics: EMNLP, 2021: 2949-2958.
[27]
LIJ, LINZ, FUP, et al. Past, present, and future: Conversational emotion recognition through structural modeling of psychological knowledge[C]//Findings of the Association for Computational Linguistics: EMNLP, 2021: 1204-1214.
[28]
PORIAS, CAMBRIAE, HAZARIKAD, et al. Context-dependent sentiment analysis in user-generated videos[C]//The 55th Annual Meeting of the Association for Computational Linguistics, 2017: 873-883.