在社交媒体中存在大量的对话文本,而在这些对话中,说话人的情感和意图通常是相关的。不仅如此,对话的整体结构也会影响对话的情感和意图,因此,需要对对话中的情感和意图进行联合学习。为此,该文提出了基于对话结构的情感、意图联合学习模型,考虑对话内潜在的情感与意图的关联性,并且利用对话的内在结构与说话人的情感和意图之间的关系,提升多轮对话文本的每一子句情感及其意图的分类性能。同时,通过使用注意力机制,利用对话的前后联系来综合考虑上下文对对话情感的影响。实验表明,联合学习模型能有效地提高对话子句情感及意图分类的性能。
Abstract
A correlation is usually exist between speaker’s sentiment and act in daily dialogs, which could also be reflected in the dialogue structure. Therefore, we propose a joint model to classify the sentiment and act in each utterance by using the dialog structure. Moreover, we use the attention mechanism to capture the impact of the structure of dialog on the sentiment of each utterance. Experiments show that the proposed model outperforms the state-of-the-art models in both dialog sentiment classification and act classification.
关键词
情感分类 /
联合学习 /
注意力机制
{{custom_keyword}} /
Key words
sentiment classication /
joint training /
attention mechanism
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Ruusuvuori J. Emotion, affect and conversation[M]. The Handbook of Conversation Analysis, 2013.
[2] Zhang L, Erickson L B, Webb H C. Effects of "emotional text" on online customer service chat[C]//Proceedings of Graduate Student Research Conference in Hospitality and Tourism, 2011:1-8.
[3] Wilson T, Wiebe J, Hoffmann P. Recognizing contextualpolarity in phrase-level sentiment analysis[J]. Proceedings of HLT/EMNLP-05, 2005, 7(5):347-354.
[4] Craggs R, Wood M M G. A categorical annotation scheme for emotion in the linguistic content of dialogue[C]//Proceedings of Tutorial and Research Workshop on Affective Dialogue Systems. Springer, Berlin, Heidelberg, 2004: 89-100.
[5] Matsumoto K, Ren F,Kuroiwa S, et al. Emotion estimation algorithm based on interpersonal emotion included in emotional dialogue sentences[C]//Proceedings of the Mexican International Conference on Artificial Intelligence. Springer, Berlin, Heidelberg, 2007: 1035-1045.
[6] Herzig J,Feigenblat G, Shmueli-Scheuer M, et al. Classifying emotions in customer support dialogues in social media[C]//Proceedings of the 17th Annual Meeting of the Special Interest Group on Discourse and Dialogue, 2016: 64-73.
[7] Cerisara C, Jafaritazehjani S, Oluokun A, et al. Multi-task dialog act and sentiment recognition on Mastodon[J]. arXiv preprint arXiv:1807.05013, 2018.
[8] Majumder N,Poria S, Peng H, et al. Sentiment and sarcasm classification with multitask learning[J]. arXiv preprint arXiv:1901.08014, 2019.
[9] Stolcke A, Ries K, Coccaro N, et al. Dialogue act modeling for automatic tagging and recognition of conversational speech[J]. Computational Linguistics, 2000, 26(3): 339-373.
[10] Lendvai P, Geertzen J. Token-based chunking of turn-internal dialogue act sequences[C]//Proceedings of the 8th SIGDIAL Workshop on Discourse and Dialogue, 2007: 174-181.
[11] Ji Y,Haffari G, Eisenstein J. A latent variable recurrent neural network for discourse relation language models[J]. arXiv preprint arXiv:1603.01913, 2016.
[12] Khanpour H, Guntakandla N, Nielsen R. Dialogue act classification in domain-independent conversations using a deep recurrent neural network[C]//Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers. 2016: 2012-2021.
[13] Can A E. Unsupervised dialogue act modeling for tutorial dialogue systems[M].Norch Carolina State Uniersity,2015.
[14] Lee J Y,Dernoncourt F. Sequential short-text classification with recurrent and convolutional neural networks[J]. arXiv preprint arXiv:1603.03821,2016.
[15] Kalchbrenner N, Blunsom P. Recurrent convolutional neural networks for discourse compositionality[J]. arXiv preprint arXiv:1306.3584, 2013.
[16] Kumar H, Agarwal A, Dasgupta R, et al. Dialogue act sequence labeling using hierarchical encoder withcrf[C]//Proceedings of 32nd AAAI Conference on Artificial Intelligence, 2018.
[17] Li Y,Su H, Shen X, et al. Dailydialog: A manually labelled multi-turn dialogue dataset[J]. arXiv preprint arXiv:1710.03957, 2017.
[18] Kim Y. Convolutional neural networks for sentence classification[J]. arXiv preprint arXiv:1408.5882, 2014.
[19] Wang Y, Feng S, Wang D, et al. Context-aware Chinese microblog sentiment classification with bidirectional LSTM[C]//Proceedings of the Asia-Pacific Web Conference. Springer, Cham, 2016: 594-606.
[20] Hazarika D,Poria S, Mihalcea R, et al. ICON: Interactive conversational memory network for multimodal emotion detection[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, 2018: 2594-2604.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金(61806137,61702518);江苏省高等学校自然科学研究面上项目(18KJB520043)
{{custom_fund}}
PDF(1129 KB)
