基于BERT 的金融文本情感分析模型

doi:10.12066/j.issn.1007-2861.2308

Abstract

Abstract:

n the financial sector, more and more investors choose to express their opinions on the internet platform. These comment texts can fully reflect investor sentiment and influence their investment decisions and market trends. Emotion analysis as an important branch of natural language processing (NLP), which provides an effective research means for analyzing a large number of text emotional types in financial sector. However, due to the professional nature of domain-specific texts and the inapplicability of large label data sets, text emotion analysis in the financial field has brought great challenges to the traditional emotion analysis model. When the general emotion analysis model is applied to specific fields such as finance, its accuracy and recall rate are poor. In order to overcome these challenges, a BERT (bidirectional encoder representations from Transformers) preprocessing model based on full word coverage and feature enhancement in financial field was proposed for the emotional analysis task of financial text from the perspective of word representation model.

Key words: sentiment analysis, word embedded vector, BERT, bag-of-POS (part of speech), named entity recognition

CLC Number:

TP 391.1

ZHU He, LU Xiaofeng, XUE Lei. Emotional analysis model of financial text based on the BERT[J]. Journal of Shanghai University（Natural Science Edition）, 2023, 29(1): 118-128.

Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks

URL: https://www.journal.shu.edu.cn/EN/10.12066/j.issn.1007-2861.2308

https://www.journal.shu.edu.cn/EN/Y2023/V29/I1/118

Figures/Tables 7

Table 1

Fig. 1

Fig. 2

Table 2

Table 3

Table 4

Table 5

[1]	Pang B, Lee L. Opinion mining and sentiment analysis[M]. Hanover: Now Publishers Inc, 2008.
[2]	Jiao J, Zhou Y. Sentiment polarity analysis based multi-dictionary[J]. Physics Procedia, 2011, 22: 590-596.
[3]	Jurek A, Mulvenna M D, Bi Y. Improved lexicon-based sentiment analysis for social media analytics[J]. Security Informatics, 2015, 4(1): 9.
[4]	Li F. The information content of forward-looking statements in corporate filings: a naÏve Bayesian machine learning approach[J]. Journal of Accounting Research, 2010, 48: 1049-1102.
[5]	Hai Z, Cong G, Chang K, et al. Analyzing sentiments in one go: a supervised joint topic modeling approach[J]. IEEE Transactions on Knowledge and Data Engineering, 2017, 29(6): 1172-1185.
[6]	Singh J, Singh G, Singh R. Optimization of sentiment analysis using machine learning classifiers[J]. Human-centric Computing and Information Sciences, 2017, 7: 1-32.
[7]	Al-Amrani Y, Lazaar M, El-Kadiri K E. Random forest and support vector machine based hybrid approach to sentiment analysis[J]. Procedia Computer Science, 2018, 127: 511-520.
[8]	杨开漠, 吴明芬, 陈涛. 广义文本情感分析综述[J]. 计算机应用, 2019, 39(S2): 6-14.
[9]	Man X, Luo T, Lin J. Financial sentiment analysis (FSA): a survey[C]// 2019 IEEE International Conference on Industrial Cyber Physical Systems (ICPS). 2019: 617-622.
[10]	Devlin J, Chang M W, Lee K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[C]// Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies. 2019: 4171-4186.
[11]	Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]// Proceedings of the 31st International Conference on Neural Information Processing Systems. 2017: 6000-6010.
[12]	Matthew E P, Neumann M, Lyyer M, et al. Deep contextualized word representa tions[C]// Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies. 2018: 2227-2237.
[13]	Alec R, Karthik N, Tim S, et al. Improving language understanding by generative pre-training[EB/OL]. [[2020-12-01].http://www.nlpir.org/wordpress/wp-content/uploads/2019/06/Improving-language-understanding-by-generative-pre-training.pdf.
[14]	Radford A, Wu J, Child R, et al. Language models are unsupervised multitask learners[EB/OL]. [[2020-12-01].https://d4mucfpksywv.cloudfront.net/better-language-odels/language-models.pdf.
[15]	Yang Z L, Dai Z H, Yang Y M, et al. XLNet: Generalized autoregressive pretraining for language understanding[C]// Proceedings of the 33rd International Conference on Neural Information Processing Systems. 2019: 5753-5763.
[16]	Dai, Z H, Yang Z L, Yang Y M, et al. Transformer-XL: attentive language models beyond a fixed-length context[C]// Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. 2019: 2978-2988.
[17]	祝清麟, 梁斌, 徐睿峰, 等. 结合金融领域情感词典和注意力机制的细粒度情感分析[J]. 中文信息学报, 2022, 36(18): 109-117.

Emotional analysis model of financial text based on the BERT

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 7

Related Articles 15

Recommended Articles

Metrics

Comments

[1]	CHEN Qian, WU Xing. Material data named entity recognition based on matching contextual lexical words and graph convolution [J]. Journal of Shanghai University（Natural Science Edition）, 2022, 28(3): 372-385.
[2]	JIN Yanliang, XIE Jinfei, WU Dijia. Chinese nested named entity recognition based on hierarchical tagging [J]. Journal of Shanghai University（Natural Science Edition）, 2022, 28(2): 270-280.
[3]	YANG Yipu, ZHU Yonghua, GAO Haiyan, GAO Wenjing. Incorporating article information for sentiment analysis of news comments [J]. Journal of Shanghai University（Natural Science Edition）, 2022, 28(1): 170-178.
[4]	ZHANG Ke, ZHANG Wenjun, ZHU Yunwen, XING Yixue. Inner-relation modelling with memory networks in aspect-based sentiment analysis [J]. Journal of Shanghai University（Natural Science Edition）, 2022, 28(1): 157-169.
[5]	WANG Xiaoxia, YUAN Xueying. Integration of Appell function and Humbert function [J]. Journal of Shanghai University（Natural Science Edition）, 2021, 27(5): 907-918.
[6]	LU Xiaohua, WANG Yongchao, DING Yang. Quasi-twisted codes achieving the Gilbert-Varshamov bound [J]. Journal of Shanghai University（Natural Science Edition）, 2021, 27(2): 289-297.
[7]	ZHOU Jingyi, GUO Yan, DING Youdong. Sentiment analysis of Chinese movie reviews based on deep learning [J]. Journal of Shanghai University（Natural Science Edition）, 2018, 24(5): 703-712.
[8]	YANG Bi-Cheng-1, CHEN Qiang-2. Half-Discrete Hilbert-Type Inequality Containing Logarithmic Kernel [J]. Journal of Shanghai University（Natural Science Edition）, 2014, 20(6): 726-732.
[9]	YANG Bi-cheng. Half-Discrete Hilbert Inequality with a Single Parameter [J]. Journal of Shanghai University（Natural Science Edition）, 2012, 18(5): 484-488.
[10]	YANG Bi-cheng. Hilbert-Type Integral Inequality with Non-homogeneous Kernel [J]. Journal of Shanghai University（Natural Science Edition）, 2011, 17(5): 603-605.
[11]	MA Ya-li,YE Wan-zhou. Perturbation of Weighted Frames in Hilbert Space and Applications [J]. Journal of Shanghai University（Natural Science Edition）, 2010, 16(3): 277-280.
[12]	YANG Bi-Cheng. A Basic Hilbert-Type Integral Inequality with Homogeneous Kernel of -1-Degree [J]. Journal of Shanghai University（Natural Science Edition）, 2009, 15(5): 493-495.
[13]	FENG Su-Xiao-1, ZHANG Dao-Xiang-1, 2, LEI Zhi-Meng-1, LIU Yu-Liu-1. D’Alembert Motions for Couple Stress Fluids [J]. Journal of Shanghai University（Natural Science Edition）, 2009, 15(3): 296-300.
[14]	YANG Bi-cheng. New Hilbert-Type Inequality [J]. Journal of Shanghai University（Natural Science Edition）, 2007, 13(3): 274-278 .
[15]	ZHONG Wu yi;YANG Bi cheng . An Equivalent Form of Extended Hardy Hilbert Integral Inequality [J]. Journal of Shanghai University（Natural Science Edition）, 2007, 13(1): 51-54 .