Multi-Granularity Semantic Aware Graph Model for Reducing Position Bias in Emotion-Cause Pair Extraction

被引:0
|
作者
Bao, Yinan [1 ,2 ]
Mao, Qianwen [1 ,2 ]
Wei, Lingwei [1 ,2 ]
Zhou, Wei [1 ]
Hu, Songlin [1 ,2 ]
机构
[1] Chinese Acad Sci, Inst Informat Engn, Beijing, Peoples R China
[2] Univ Chinese Acad Sci, Sch Cyber Secur, Beijing, Peoples R China
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The Emotion-Cause Pair Extraction (ECPE) task aims to extract emotions and causes as pairs from documents. We observe that the relative distance distribution of emotions and causes is extremely imbalanced in the typical ECPE dataset. Existing methods have set a fixed size window to capture relations between neighboring clauses. However, they neglect the effective semantic connections between distant clauses, leading to poor generalization ability towards position-insensitive data. To alleviate the problem, we propose a novel Multi-Granularity Semantic Aware Graph model (MGSAG) to incorporate fine-grained and coarse-grained semantic features jointly, without regard to distance limitation. In particular, we first explore semantic dependencies between clauses and keywords extracted from the document that convey fine-grained semantic features, obtaining keywords enhanced clause representations. Besides, a clause graph is also established to model coarse-grained semantic relations between clauses. Experimental results indicate that MGSAG surpasses the existing state-of-the-art ECPE models. Especially, MGSAG outperforms other models significantly in the condition of position-insensitive data.
引用
收藏
页码:1203 / 1213
页数:11
相关论文
共 49 条
  • [11] A graph attention network utilizing multi-granular information for emotion-cause pair extraction
    Chen, Siyuan
    Mao, Kezhi
    NEUROCOMPUTING, 2023, 543
  • [12] Emotion-cause pair extraction via knowledge-driven multi-classification and graph-based position embedding
    Linlin Zong
    Jinglin Zhang
    Jiahui Zhou
    Xianchao Zhang
    Bo Xu
    Applied Intelligence, 2024, 54 : 2703 - 2715
  • [13] Multi-granularity semantic representation model for relation extraction
    Ming Lei
    Heyan Huang
    Chong Feng
    Neural Computing and Applications, 2021, 33 : 6879 - 6889
  • [14] Emotion-cause pair extraction via knowledge-driven multi-classification and graph-based position embedding
    Zong, Linlin
    Zhang, Jinglin
    Zhou, Jiahui
    Zhang, Xianchao
    Xu, Bo
    APPLIED INTELLIGENCE, 2024, 54 (03) : 2703 - 2715
  • [15] Co-Evolving Graph Reasoning Network for Emotion-Cause Pair Extraction
    Xing, Bowen
    Tsang, Ivor W.
    MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES: RESEARCH TRACK, ECML PKDD 2023, PT I, 2023, 14169 : 305 - 322
  • [16] A knowledge-guided graph attention network for emotion-cause pair extraction
    Zhu, Peican
    Wang, Botao
    Tang, Keke
    Zhang, Haifeng
    Cui, Xiaodong
    Wang, Zhen
    KNOWLEDGE-BASED SYSTEMS, 2024, 286
  • [17] Transition-based Directed Graph Construction for Emotion-Cause Pair Extraction
    Fan, Chuang
    Yuan, Chaofa
    Du, Jiachen
    Gui, Lin
    Yang, Min
    Xu, Ruifeng
    58TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2020), 2020, : 3707 - 3717
  • [18] A knowledge-guided graph attention network for emotion-cause pair extraction
    Zhu, Peican
    Wang, Botao
    Tang, Keke
    Zhang, Haifeng
    Cui, Xiaodong
    Wang, Zhen
    Knowledge-Based Systems, 2024, 286
  • [19] Emotion-cause pair extraction based on machine reading comprehension model
    Chang, Ting Wei
    Fan, Yao-Chung
    Chen, Arbee L. P.
    MULTIMEDIA TOOLS AND APPLICATIONS, 2022, 81 (28) : 40653 - 40673
  • [20] Emotion-cause pair extraction based on machine reading comprehension model
    Chang, Ting Wei
    Fan, Yao-Chung
    Chen, Arbee L.P.
    Multimedia Tools and Applications, 2022, 81 (28): : 40653 - 40673