Please wait a minute...
Data Analysis and Knowledge Discovery  2022, Vol. 6 Issue (7): 118-127    DOI: 10.11925/infotech.2096-3467.2021.1344
Original article Current Issue | Archive | Adv Search |
Identifying Financial Text Causality with Bi-LSTM and Two-way CNN
Zhang Shunxiang(),Zhang Zhenjiang,Zhu Guangli,Zhao Tong,Huang Ju
School of Computer Science and Engineering, Anhui University of Science & Technology, Huainan 232001, China
Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei 230088, China
Download: PDF (1487 KB)   HTML ( 12
Export: BibTeX | EndNote (RIS)      
Abstract  

[Objective] This paper proposes a network model with Bi-LSTM and two-way CNN, which addresses the missing characteristic information for causality identification and improves its accuracy. [Methods] First, we used the Bi-LSTM to generate the text feature matrix for the financial texts. Then, we extracted the causal features from the matrix using two-way CNN with different convolution cores. Third, we spliced the feature vectors obtained by maximum and average pooling methods. Finally, we transferred the spliced vectors to the full connection layer for output. [Results] The accuracy of our new model reached 82.3%, which is at least 3% higher than those of the existing methods. [Limitations] We did not establish specific function module for the financial texts. [Conclusions] The proposed model could effectively identify the causality from the documents.

Key wordsFinancial Text      Causal Recognition      Bi-LSTM      Two-way CNN     
Received: 26 November 2021      Published: 24 August 2022
ZTFLH:  TP393 G250  
Fund:National Natural Science Foundation of China(62076006);University Synergy Innovation Program of Anhui Province(GXXT-2021-008);Anhui Provincial Key R&D Program(202004b11020029)
Corresponding Authors: Zhang Shunxiang,ORCID: 0000-0002-0540-7593     E-mail: sxzhang@aust.edu.cn

Cite this article:

Zhang Shunxiang, Zhang Zhenjiang, Zhu Guangli, Zhao Tong, Huang Ju. Identifying Financial Text Causality with Bi-LSTM and Two-way CNN. Data Analysis and Knowledge Discovery, 2022, 6(7): 118-127.

URL:

https://manu44.magtech.com.cn/Jwk_infotech_wk3/EN/10.11925/infotech.2096-3467.2021.1344     OR     https://manu44.magtech.com.cn/Jwk_infotech_wk3/EN/Y2022/V6/I7/118

Recognition of Causality in Financial Text
LSTM Structure
Acquire Text Feature Matrix
Single Convolutional Neural Network
Model Iteration Process
数据集 数量 百分比/%
OIEC 6 824 74.4
AA 2 348 25.6
OIECAA Dataset Distribution
数据集 集合 句子数 句子最大长度
OIEC 训练集 5 459 128
验证集 1 365
测试集 1 365
AA 训练集 1 878 117
验证集 470
测试集 470
OIECAA Data Set Partition
参数
embedding_size 128
learning_rate 0.2
Epoch 70
batch_size 128
dropout 0.6
CNN_filter 6
Parameter Settings
实验 P/% R/% F1/%
LSTM+CNN 75.37 77.19 76.27
Bi-LSTM+CNN 78.97 79.72 79.34
LSTM+双路CNN 77.86 78.47 78.16
BTCNN 82.30 80.04 81.15
Test Results
Iterative Training Results
[1] 李朋远, 于华, 江成. 中文语境下基于事件关联挖掘的金融网络构建与分析[J]. 中国科学院大学学报, 2021, 38(2): 270-279.
doi: 10.7523/j.issn.2095-6134.2021.02.013
[1] ( Li Pengyuan, Yu Hua, Jiang Cheng. Financial Network Construction and Analysis Based on Event Correlation Extraction under Chinese Context[J]. Journal of University of Chinese Academy of Sciences, 2021, 38(2): 270-279.)
doi: 10.7523/j.issn.2095-6134.2021.02.013
[2] 单晓红, 庞世红, 刘晓燕, 等. 基于事理图谱的网络舆情事件预测方法研究[J]. 情报理论与实践, 2020, 43(10): 165-170.
[2] ( Shan Xiaohong, Pang Shihong, Liu Xiaoyan, et al. Research on Internet Public Opinion Event Prediction Method Based on Event Evolution Graph[J]. Information Studies: Theory & Application, 2020, 43(10): 165-170.)
[3] 刘炜, 王旭, 张雨嘉, 等. 一种面向突发事件的文本语料自动标注方法[J]. 中文信息学报, 2017, 31(2): 76-85.
[3] ( Liu Wei, Wang Xu, Zhang Yujia, et al. An Automatic-Annotation Method for Emergency Text Corpus[J]. Journal of Chinese Information Processing, 2017, 31(2): 76-85.)
[4] 徐荣, 郭娜, 李金鑫, 等. 我国房地产价格波动对系统性金融风险影响的动态机制研究——基于有向无环图的分析[J]. 南方经济, 2017(11): 1-17.
[4] ( Xu Rong, Guo Na, Li Jinxin. Study on the Impact of Real Estate Price Fluctuation on Financial System Risk in China: Based on Directed Acyclic Graph Analysis[J]. South China Journal of Economics, 2017(11): 1-17.)
[5] Blanco E, Castell N, Moldovan D. Causal Relation Extraction[C]// Proceedings of the 6th International Conference on Language Resources and Evaluation. 2008: 310-313.
[6] 马彬, 洪宇, 杨雪蓉, 等. 基于语义依存线索的事件关系识别方法研究[J]. 北京大学学报(自然科学版), 2013, 49(1): 109-116.
[6] ( Ma Bin, Hong Yu, Yang Xuerong, et al. Using Event Dependency Cue Inference to Recognize Event Relation[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2013, 49(1): 109-116.)
[7] Girju R. Automatic Detection of Causal Relations for Question Answering[C]// Proceedings of the ACL 2003 Workshop on Multilingual Summarization and Question Answering. 2003: 76-83.
[8] 王军, 夏利民. 基于因果分析的群体行为识别[J]. 电子科技大学学报, 2018, 47(2): 256-261.
[8] ( Wang Jun, Xia Limin. Group Activity Recognition in Crowd via Causality Analysis[J]. Journal of University of Electronic Science and Technology of China, 2018, 47(2): 256-261.)
[9] 郑忠国, 张艳艳, 童行伟. 因果模型中因果效应的可识别性研究[J]. 中国科学: 数学A辑, 2001, 31(12): 1080-1086.
[9] Zheng Zhongguo, Zhang Yanyan, Tong Xingwei. Research on the Recognizability of Causal Effect in Causal Model[J]. Science in China (Series A), 2001, 31(12): 1080-1086.)
[10] 郑新, 李培峰, 朱巧明. 中文事件时序关系的标注和分类方法[J]. 计算机科学, 2015, 42(7): 276-279, 313.
[10] ( Zheng Xin, Li Peifeng, Zhu Qiaoming. Annotation and Classification of Temporal Relation Between Chinese Events[J]. Computer Science, 2015, 42(7): 276-279, 313.)
[11] 赵亮, 方芳, 王伟, 等. 基于脑电信号间Granger因果关系特征的情感识别[J]. 电子测量与仪器学报, 2018, 32(6): 87-95.
[11] ( Zhao Liang, Wang Wei, et al. Emotion Recognition Based on Granger Causality Feature Between EEG Signals[J]. Journal of Electronic Measurement and Instrumentation, 2018, 32(6): 87-95.)
[12] Garcia D. COATIS, an NLP System to Locate Expressions of Actions Connected by Causality Links[C]// Proceedings of International Conference on Knowledge Acquisition, Modeling and Management. 1997: 347-352.
[13] 杜小坤, 李国徽, 李艳红. 模式匹配中的结构差异识别及消解[J]. 计算机科学, 2015, 42(2): 185-190.
[13] ( Du Xiaokun, Li Guohui, Li Yanhong. Structural Difference Recognition and Dispelling in Schema Matching[J]. Computer Science, 2015, 42(2): 185-190 )
[14] 孙佳伟, 李正华, 陈文亮, 等. 基于词模式嵌入的词语上下位关系分类[J]. 北京大学学报(自然科学版), 2019, 55(1): 1-7.
[14] ( Sun Jiawei, Li Zhenghua, Chen Wenliang, et al. Hypernym Relation Classification Based on Word Pattern[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2019, 55(1): 1-7.)
[15] 杜小坤, 李国徽, 王江晴, 等. 基于信息元的模式匹配方法[J]. 软件学报, 2015, 26(10): 2596-2613.
[15] ( Du Xiaokun, Li Guohui, Wang Jiangqing, et al. Schema Matching Method Based on Information Unit[J]. Journal of Software, 2015, 26(10): 2596-2613.)
[16] 马彬, 洪宇, 杨雪蓉, 等. 基于语义依存线索的事件关系识别方法研究[J]. 北京大学学报(自然科学版), 2013, 49(1): 109-116.
[16] ( Ma Bin, Hong Yu, Yang Xuerong, et al. Using Event Dependency Cue Inference to Recognize Event Relation[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2013, 49(1): 109-116.)
[17] Nauta M, Bucur D, Seifert C. Causal Discovery with Attention-based Convolutional Neural Networks[J]. Machine Learning and Knowledge Extraction, 2019, 1(1): 312-340.
doi: 10.3390/make1010019
[18] 李培峰, 黄一龙, 朱巧明. 使用全局优化方法识别中文事件因果关系[J]. 清华大学学报(自然科学版), 2017, 57(10): 1042-1047.
[18] Li Peifeng, Huang Yilong, Zhu Qiaoming. Global Optimization to Recognize Causal Relations Between Events[J]. Journal of Tsinghua University (Science and Technology), 2017, 57(10): 1042-1047.)
[19] 杨竣辉, 刘宗田, 刘炜, 等. 基于语义事件因果关系识别[J]. 小型微型计算机系统, 2016, 37(3): 433-437.
[19] ( Yang Junhui, Liu Zongtian, Liu Wei, et al. Identify Causality Relationships Based on Semantic Event[J]. Journal of Chinese Computer Systems, 2016, 37(3): 433-437.)
[20] 黄一龙, 李培峰, 朱巧明. 事件因果与时序关系识别的联合推理模型[J]. 计算机科学, 2018, 45(6): 204-207, 234.
[20] ( Huang Yilong, Li Peifeng, Zhu Qiaoming. Joint Model of Events' Causal and Temporal Relations Identification[J]. Computer Science, 2018, 45(6): 204-207, 234.)
[21] 冯冲, 康丽琪, 石戈, 等. 融合对抗学习的因果关系抽取[J]. 自动化学报, 2018, 44(5): 811-818.
[21] ( Feng Chong, Kang Liqi, Shi Ge, et al. Causality Extraction with GAN[J]. Acta Automatica Sinica, 2018, 44(5): 811-818.)
[22] 付剑锋, 刘宗田, 刘炜, 等. 基于层叠条件随机场的事件因果关系抽取[J]. 模式识别与人工智能, 2011, 24(4): 567-573.
[22] ( Fu Jianfeng, Liu Zongtian, Liu Wei, et al. Event Causal Relation Extraction Based on Cascaded Conditional Random Fields[J]. Pattern Recognition and Artificial Intelligence, 2011, 24(4): 567-573.)
[23] Zhang S, Zheng D Q, Hu X C, et al. Bidirectional Long Short-Term Memory Networks for Relation Classification[C]// Proceedings of the 29th Pacific Asia Conference on Language, Information and Computation. 2015: 73-78.
[24] 田生伟, 周兴发, 禹龙, 等. 基于双向LSTM的维吾尔语事件因果关系抽取[J]. 电子与信息学报, 2018, 40(1): 200-208.
[24] ( Tian Shengwei, Zhou Xingfa, Yu Long, et al. Causal Relation Extraction of Uyghur Events Based on Bidirectional Long Short-Term Memory Model[J]. Journal of Electronics & Information Technology, 2018, 40(1): 200-208.)
[25] 闻畅, 刘宇, 顾进广. 基于注意力机制的双向长短时记忆网络模型突发事件演化关系抽取[J]. 计算机应用, 2019, 39(6): 1646-1651.
doi: 10.11772/j.issn.1001-9081.2018122533
[25] ( Wen Chang, Liu Yu, Gu Jinguang. Evolution Relationship Extraction of Emergency Based on Attention-Based Bidirectional Long Short-Term Memory Network Model[J]. Journal of Computer Applications, 2019, 39(6): 1646-1651.)
doi: 10.11772/j.issn.1001-9081.2018122533
[26] 马建红, 郝亚娟, 张亚梅. 基于层叠跳跃链条件随机场模型的因果关系标注[J]. 郑州大学学报(理学版), 2016, 48(4): 54-59.
[26] Ma Jianhong, Hao Yajuan, Zhang Yamei. Causal Relation Label Based on Cascading Skip-Chain Conditional Random Fields[J]. Journal of Zhengzhou University (Natural Science Edition), 2016, 48(4): 54-59.)
[27] 林建莹, 黄登笑, 盛红彬, 等. 药品不良反应因果关系判定研究[J]. 上海交通大学学报(医学版), 2010, 30(8): 951-955.
[27] Lin Jianying, Huang Dengxiao, Sheng Hongbin, et al. Study on Causality Assessment of Adverse Drug Reactions[J]. Journal of Shanghai Jiao Tong University (Medical Science), 2010, 30(8): 951-955.)
[28] 王元龙, 李茹, 张虎, 等. 阅读理解中因果关系类选项的研究[J]. 清华大学学报(自然科学版), 2018, 58(3): 272-278.
[28] Wang Yuanlong, Li Ru, Zhang Hu, et al. Causal Options in Chinese Reading Comprehension[J]. Journal of Tsinghua University (Science and Technology), 2018, 58(3): 272-278.)
[29] Ding R X, Li Z J. Event Extraction with Deep Contextualized Word Representation and Multi-attention Layer[C]// Proceedings of International Conference on Advanced Data Mining and Applications. 2018: 189-201.
[30] Yang Y L, Tong Y H, Ma S L, et al. A Position Encoding Convolutional Neural Network Based on Dependency Tree for Relation Classification[C]// Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. 2016: 65-74.
[31] 许晶航, 左万利, 梁世宁, 等. 基于图注意力网络的因果关系抽取[J]. 计算机研究与发展, 2020, 57(1): 159-174.
[31] ( Xu Jinghang, Zuo Wanli, Liang Shining, et al. Causal Relation Extraction Based on Graph Attention Networks[J]. Journal of Computer Research and Development, 2020, 57(1): 159-174.)
[1] Hu Haotian,Ji Jinfeng,Wang Dongbo,Deng Sanhong. An Integrated Platform for Food Safety Incident Entities Based on Deep Learning[J]. 数据分析与知识发现, 2021, 5(3): 12-24.
[2] Ma Jianxia,Yuan Hui,Jiang Xiang. Extracting Name Entities from Ecological Restoration Literature with Bi-LSTM+CRF[J]. 数据分析与知识发现, 2020, 4(2/3): 78-88.
[3] Qiang Lu,Zhenfang Zhu,Fuyong Xu,Qiangqiang Guo. Chinese Sentiment Classification Method with Bi-LSTM and Grammar Rules[J]. 数据分析与知识发现, 2019, 3(11): 99-107.
[4] Lianjie Xiao,Tao Meng,Wei Wang,Zhixiang Wu. Entity Recognition of Intelligence Method Based on Deep Learning: Taking Area of Security Intelligence for Example[J]. 数据分析与知识发现, 2019, 3(10): 20-28.
[5] Feng Guoming,Zhang Xiaodong,Liu Suhui. DBLC Model for Word Segmentation Based on Autonomous Learning[J]. 数据分析与知识发现, 2018, 2(5): 40-47.
  Copyright © 2016 Data Analysis and Knowledge Discovery   Tel/Fax:(010)82626611-6626,82624938   E-mail:jishu@mail.las.ac.cn