{"type":"ScholarlyArticle","title":"Conversational Semantic Role Labeling with Predicate-Oriented Latent Graph","authors":[{"name":"Hao Fei","slug":"hao-fei","given":"Hao","family":"Fei","affiliation":"Key Laboratory of Aerospace Information Security and Trusted Computing , Ministry of Education , School of Cyber Science and Engineering , Wuhan University , China"},{"name":"Shengqiong Wu","slug":"shengqiong-wu","given":"Shengqiong","family":"Wu","affiliation":"Key Laboratory of Aerospace Information Security and Trusted Computing , Ministry of Education , School of Cyber Science and Engineering , Wuhan University , China"},{"name":"Meishan Zhang","slug":"meishan-zhang","given":"Meishan","family":"Zhang","affiliation":"Institute of Computing and Intelligence , Harbin Institute of Technology (Shenzhen) , China"},{"name":"Yafeng Ren","slug":"yafeng-ren","given":"Yafeng","family":"Ren","affiliation":"Laboratory of Language and Artificial Intelligence , Guangdong University of Foreign Studies , China"},{"name":"Donghong Ji","slug":"donghong-ji","given":"Donghong","family":"Ji","affiliation":"Key Laboratory of Aerospace Information Security and Trusted Computing , Ministry of Education , School of Cyber Science and Engineering , Wuhan University , China"}],"abstract":"Conversational semantic role labeling (CSRL) is a newly proposed task that uncovers the shallow semantic structures in a dialogue text. Unfortunately several important characteristics of the CSRL task have been overlooked by the existing works, such as the structural information integration, near-neighbor influence. In this work, we investigate the integration of a latent graph for CSRL. We propose to automatically induce a predicate-oriented latent graph (POLar) with a predicate-centered gaussian mechanism, by which the nearer and informative words to the predicate will be allocated with more attention. The POLar structure is then dynamically pruned and refined so as to best fit the task need. We additionally introduce an effective dialogue-level pre-trained language model, CoDiaBERT, for better supporting multiple utterance sentences and handling the speaker coreference issue in CSRL. Our system outperforms best-performing baselines on three benchmark CSRL datasets with big margins, especially achieving over 4% F1 score improvements on the cross-utterance argument detection. Further analyses are presented to better understand the effectiveness of our proposed methods.","doi":"10.24963/ijcai.2022/571","url":"https://ijcai.org/proceedings/2022/571","pdf":"https://ijcai.org/proceedings/2022/0571.pdf","markdown":"https://ijcai.org/proceedings/2022/571.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/571","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"Natural Language Processing","track":"Main Track","number":571,"pages":{"first":4114,"last":4120},"keywords":["Natural Language Processing: Information Extraction","Natural Language Processing: Dialogue and Interactive Systems","Natural Language Processing: Tagging, Chunking, and Parsing"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38986579","length_sec":76,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38984722","length_sec":325,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Bastings et al., 2019] Jasmijn Bastings, Wilker Aziz, and Ivan Titov. Interpretable neural predictions with differen- tiable binary variables. In Proc. of ACL, pages 2963-2977, 2019.","title":"Interpretable neural predictions with differentiable binary variables","authors":"Jasmijn Bastings, Wilker Aziz, Ivan Titov","year":"2019","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Chen et al., 2020] Chenhua Chen, Zhiyang Teng, and Yue Zhang. Inducing target-specific latent structures for aspect sentiment classification. In Proc. of EMNLP, pages 5596- 5607, 2020.","title":"Inducing target-specific latent structures for aspect sentiment classification","authors":"Chen","year":"2020","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Choi et al., 2018] Jihun Choi, Kang Min Yoo, and Sang-goo Lee. Learning to compose task-specific tree structures. In Proc. of AAAI, pages 5094-5101, 2018.","title":"Learning to compose task-specific tree structures","authors":"Jihun Choi, Min Kang, Sang-Goo Yoo, Lee","year":"2018","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Correia et al., 2019] Gonc ¸alo M. Correia, Vlad Niculae, and André F. T. Martins. Adaptively sparse transformers. In Proc. of EMNLP, pages 2174-2184, 2019.","title":"Adaptively sparse transformers","authors":"Correia","year":"2019","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Corro and Titov, 2019] Caio Corro and Ivan Titov. Learn- ing latent trees with stochastic perturbations and differ- entiable dynamic programming. In Proc. of ACL, pages 5508-5521, 2019.","title":"Learning latent trees with stochastic perturbations and differentiable dynamic programming","authors":"Titov ; ] Caio Corro, Ivan Corro, Titov","year":"2019","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Devlin et al., 2019] Jacob Devlin, Ming-Wei Chang, Ken- ton Lee, and Kristina Toutanova. BERT: pre-training of deep bidirectional transformers for language understand- ing. In Proc. of NAACL, pages 4171-4186, 2019.","title":"BERT: pre-training of deep bidirectional transformers for language understanding","authors":"Jacob Devlin, Ming-Wei Chang, Kenton Lee, Kristina Toutanova","year":"2019","venue":"Proc. of NAACL","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Fei et al., 2020a] Hao Fei, Yafeng Ren, and Donghong Ji. Improving text understanding via deep syntax-semantics communication. In Findings of the Association for Com- putational Linguistics: EMNLP, pages 84-93, 2020.","title":"Improving text understanding via deep syntax-semantics communication","authors":"Fei","year":"2020","venue":"Findings of the Association for Computational Linguistics: EMNLP","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Fei et al., 2020b] Hao Fei, Yafeng Ren, and Donghong Ji. Mimic and conquer: Heterogeneous tree structure distil- lation for syntactic NLP. In Findings of the Association for Computational Linguistics: EMNLP, pages 183-193, 2020.","title":"Mimic and conquer: Heterogeneous tree structure distillation for syntactic NLP","authors":"Fei","year":"2020","venue":"Findings of the Association for Computational Linguistics: EMNLP","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Fei et al., 2020c] Hao Fei, Yafeng Ren, and Donghong Ji. Retrofitting structure-aware transformer language model for end tasks. In Proc. of EMNLP, pages 2151-2161, 2020.","title":"Retrofitting structure-aware transformer language model for end tasks","authors":"Fei","year":"2020","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Fei et al., 2020d] Hao Fei, Meishan Zhang, and Donghong Ji. Cross-lingual semantic role labeling with high-quality translated training corpus. In Proc. of AAAI, pages 7014- 7026, 2020.","title":"Cross-lingual semantic role labeling with high-quality translated training corpus","authors":"Fei","year":"2020","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Fei et al., 2021a] Hao Fei, Fei Li, Bobo Li, and Donghong Ji. Encoder-decoder based unified semantic role labeling with label-aware syntax. In Proc. of AAAI, pages 12794- 12802, 2021.","title":"Encoder-decoder based unified semantic role labeling with label-aware syntax","authors":"Fei","year":"2021","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Fei et al., 2021b] Hao Fei, Shengqiong Wu, Yafeng Ren, Fei Li, and Donghong Ji. Better combine them together! integrating syntactic constituency and dependency repre- sentations for semantic role labeling. In Findings of the Association for Computational Linguistics: ACL/IJCNLP 2021, pages 549-559, 2021.","title":"Better combine them together! integrating syntactic constituency and dependency representations for semantic role labeling","authors":"Fei","year":"2021","venue":"Findings of the Association for Computational Linguistics: ACL/IJCNLP 2021","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Fei et al., 2021c] Hao Fei, Meishan Zhang, Bobo Li, and Donghong Ji. End-to-end semantic role labeling with neu- ral transition-based model. In Proc. of AAAI, pages 12803- 12811, 2021.","title":"End-to-end semantic role labeling with neural transition-based model","authors":"Fei","year":"2021","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Gildea and Jurafsky, 2000] Daniel Gildea and Daniel Juraf- sky. Automatic labeling of semantic roles. In Proc. of ACL, pages 512-520, 2000.","title":"Automatic labeling of semantic roles","authors":"Jurafsky Gildea","year":"2000","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[He et al., 2018] Shexia He, Zuchao Li, Hai Zhao, and Hongxiao Bai. Syntax for semantic role labeling, to be, or not to be. In Proc. of ACL, pages 2061-2071, 2018.","title":"Syntax for semantic role labeling, to be, or not to be","authors":"He","year":"2018","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Kumaraswamy, 1980] Ponnambalam Kumaraswamy. A generalized probability density function for double- bounded random processes. Journal of Hydrology, 46(1- 2):79-88, 1980.","title":"A generalized probability density function for doublebounded random processes","authors":"Ponnambalam Kumaraswamy","year":"1980","venue":"Journal of Hydrology","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Li et al., 2019] Zuchao Li, Shexia He, Hai Zhao, Yiqing Zhang, Zhuosheng Zhang, Xi Zhou, and Xiang Zhou. De- pendency or span, end-to-end uniform semantic role label- ing. In Proc. of AAAI, pages 6730-6737, 2019.","title":"Dependency or span, end-to-end uniform semantic role labeling","authors":"Li","year":"2019","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Li et al., 2020] Jingye Li, Hao Fei, and Donghong Ji. Mod- eling local contexts for joint dialogue act recognition and sentiment classification with bi-channel dynamic convolu- tions. In Proc. of COLING, pages 616-626, 2020.","title":"Modeling local contexts for joint dialogue act recognition and sentiment classification with bi-channel dynamic convolutions","authors":"Li","year":"2020","venue":"Proc. of COLING","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Liu and Gildea, 2010] Ding Liu and Daniel Gildea. Seman- tic role features for machine translation. In Proc. of COL- ING, pages 716-724, 2010.","title":"Semantic role features for machine translation","authors":"Gildea ; ] Ding Liu, Daniel Liu, Gildea","year":"2010","venue":"Proc. of COL-ING","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Liu and Lapata, 2019] Yang Liu and Mirella Lapata. Text summarization with pretrained encoders. In Proc. of EMNLP, pages 3730-3740, 2019.","title":"Text summarization with pretrained encoders","authors":"Yang Liu, Mirella Lapata","year":"2019","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Marcheggiani and Titov, 2017] Diego Marcheggiani and Ivan Titov. Encoding sentences with graph convolutional networks for semantic role labeling. In Proc. of EMNLP, pages 1506-1515, 2017.","title":"Diego Marcheggiani and Ivan Titov. Encoding sentences with graph convolutional networks for semantic role labeling","authors":"Titov Marcheggiani","year":"2017","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Shen and Lapata, 2007] Dan Shen and Mirella Lapata. Us- ing semantic roles to improve question answering. In Proc. of EMNLP, pages 12-21, 2007.","title":"Using semantic roles to improve question answering","authors":"Lapata ; Shen, Dan Shen, Mirella Lapata","year":"2007","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Shi and Lin, 2019] Peng Shi and Jimmy Lin. Simple BERT models for relation extraction and semantic role labeling. CoRR, abs/1904.05255, 2019.","title":"Simple BERT models for relation extraction and semantic role labeling","authors":"Peng Shi, Jimmy Lin","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Strubell et al., 2018] Emma Strubell, Patrick Verga, Daniel Andor, David Weiss, and Andrew McCallum. Linguistically-informed self-attention for semantic role labeling. In Proc. of EMNLP, pages 5027-5038, 2018.","title":"Linguistically-informed self-attention for semantic role labeling","authors":"Emma Strubell, Patrick Verga, Daniel Andor, David Weiss, Andrew Mccallum","year":"2018","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Wang et al., 2015] Hai Wang, Mohit Bansal, Kevin Gimpel, and David McAllester. Machine comprehension with syn- tax, frames, and semantics. In Proc. of ACL, pages 700- 706, 2015.","title":"Machine comprehension with syntax, frames, and semantics","authors":"Hai Wang, Mohit Bansal, Kevin Gimpel, David Mcallester","year":"2015","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Wu et al., 2021a] Han Wu, Kun Xu, Linfeng Song, Lifeng Jin, Haisong Zhang, and Linqi Song. Domain-adaptive pretraining methods for dialogue understanding. In Proc. of ACL, pages 665-669, 2021.","title":"Domain-adaptive pretraining methods for dialogue understanding","authors":"Wu","year":"2021","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Wu et al., 2021b] Han Wu, Kun Xu, and Linqi Song. CSAGN: Conversational structure aware graph network for conversational semantic role labeling. In Proc. of EMNLP, pages 2312-2317, 2021.","title":"CSAGN: Conversational structure aware graph network for conversational semantic role labeling","authors":"Wu","year":"2021","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Xu et al., 2021] Kun Xu, Han Wu, Linfeng Song, Haisong Zhang, Linqi Song, and Dong Yu. Conversational seman- tic role labeling. IEEE ACM Trans. Audio Speech Lang. Process., 29:2465-2475, 2021.","title":"Conversational semantic role labeling","authors":"Kun Xu, Han Wu, Linfeng Song, Haisong Zhang, Linqi Song, Dong Yu","year":"2021","venue":"IEEE ACM Trans. Audio Speech Lang. Process","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}