[1]段魏诚,薛涛.FinBERT-RCNN-ATTACK:金融文本情感分析模型[J].计算机技术与发展,2024,34(05):157-162.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0054]
　DUAN Wei-cheng,XUE Tao.FinBERT-RCNN-ATTACK:Emotional Analysis Model of Financial Text[J].,2024,34(05):157-162.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0054]
点击复制

FinBERT-RCNN-ATTACK:金融文本情感分析模型()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
34
期数:
2024年05期
页码:
157-162
栏目:
人工智能
出版日期:
2024-05-10

文章信息/Info

Title:
FinBERT-RCNN-ATTACK:Emotional Analysis Model of Financial Text
文章编号:
1673-629X(2024)05-0157-06
作者:
段魏诚; 薛涛
西安工程大学 计算机科学学院,陕西 西安 710600
Author(s):
DUAN Wei-cheng; XUE Tao
School of Computer Science,Xi’an Polytechnic University,Xi’an 710600,China
关键词:
金融文本; 情感分析; FinBERT; 循环卷积神经网络; 对抗训练
Keywords:
financial text; emotional analysis; FinBERT; recurrent convolutional neural networks; adversarial training
分类号:
TP391
DOI:
10.20165/j.cnki.ISSN1673-629X.2024.0054
摘要:
金融文本包含投资者的情绪及公众对相关事件的态度。 近年来,自然语言处理已广泛应用于金融领域,对金融文本数据进行情感分析可以得到丰富的投资价值和监管参考价值。 然而由于金融词汇具有专业性和特殊性,现有的通用情感分析模型不适合金融领域情感分析任务,精确度有待提高,且现有模型易受到对抗样本的干扰导致模型结果出错。 为了解决这些问题,提出了一个 FinBERT-RCNN-ATTACK 模型。 利用在金融语料库预训练的 FinBERT 模型进行词嵌入处理,提取语义特征,将提取到的特征引入 RCNN 模型进一步挖掘上下文的关键特征,并且在模型中引入对抗训练,即在嵌入阶段添加扰动,提高模型的鲁棒性和泛化性。 实验结果表明,在金融领域数据集上,提出的模型优于其他情感分析模型,精准度提升了 3% ~ 35% 。
Abstract:
The financial text contains investor sentiment and public attitudes towards the events. In recent years, natural language processing has been widely used in financial field,and the emotional analysis of financial text data can get rich investment value and regu-latory reference value. However, due to the professionalism and particularity of financial vocabulary, the existing general emotional analysis model is not suitable for the emotional analysis task in the financial field,and the accuracy needs to be improved,and the existing model is vulnerable to the interference of antagonistic samples,leading to the wrong model results. In order to solve these problems,we proposed a FinBERT-RCNN-ATTACK model. The FinBERT model pre-trained in the financial corpus is used for word embedding processing to extract semantic features,and the extracted features are introduced into the RCNN model to further excavate the key features of the context. In addition, adversarial training is introduced into the model, that is, disturbance is added in the embedding stage toimprove the robustness and generalization of the model. The experimental results show that the proposed model is better than the other e-motional analysis models,and the accuracy is improved by 3% ~ 35% .

相似文献/References:

[1]李妍坊,许歆艺,刘功申. 面向情感倾向性识别的特征分析研究[J].计算机技术与发展,2014,24(09):33.
　LI Yan-fang,XU Xin-yi,LIU Gong-shen. Research on Feature Analysis Oriented Text Sentiment Identification[J].,2014,24(05):33.
[2]苏小英[][],孟环建[]. 基于神经网络的微博情感分析[J].计算机技术与发展,2015,25(12):161.
　SU Xiao-ying[][],MENG Huan-jian[]. Sentiment Analysis of Micro-blog Based on Neural Networks[J].,2015,25(05):161.
[3]陈耀东,彭蝶飞.一种面向旅游评论的情感特征识别方法[J].计算机技术与发展,2018,28(11):107.[doi:10.3969/ j. issn.1673-629X.2018.11.24]
　CHEN Yao-dong,PENG Die-fei.A Recognition Method of Sentiment Features for Tour Reviews[J].,2018,28(05):107.[doi:10.3969/ j. issn.1673-629X.2018.11.24]
[4]杨立月,王移芝.微博情感分析的情感词典构造及分析方法研究[J].计算机技术与发展,2019,29(02):13.[doi:10．3969/j．issn．1673－629X．2019．02．003]
　YANG Liyue,WANG Yizhi.Research on Construction and Analysis of Emotion Dictionary in Emotion Analysis of Micro－blog[J].,2019,29(05):13.[doi:10．3969/j．issn．1673－629X．2019．02．003]
[5]唐 莉,刘 臣.基于 CRF 和 HITS 算法的特征情感对提取[J].计算机技术与发展,2019,29(07):71.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 014]
　TANG Li,LIU Chen.Extraction of Feature and Sentiment Word Pair Based on Conditional Random Fields and HITS Algorithm[J].,2019,29(05):71.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 014]
[6]申静波,李井辉,孙丽娜.注意力机制在评论文本情感分析中的应用研究[J].计算机技术与发展,2020,30(07):169.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 036]
　SHEN Jing-bo,LI Jing-hui,SUN Li-na.Research on Application of Attention Mechanism in Comment Text Emotional Analysis[J].,2020,30(05):169.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 036]
[7]邱全磊,崔宗敏,喻 静.基于表情和语气的情感词典用于弹幕情感分析[J].计算机技术与发展,2020,30(08):178.[doi:10. 3969 / j. issn. 1673-629X. 2020. 08. 031]
　QIU Quan-lei,CUI Zong-min,YU Jing.Emotional Dictionary Based on Emoticons and Modal for Barrage Sentiment Analysis[J].,2020,30(05):178.[doi:10. 3969 / j. issn. 1673-629X. 2020. 08. 031]
[8]王连喜.基于“属性-情感词”汽车本体的文本情感分析[J].计算机技术与发展,2020,30(08):193.[doi:10. 3969 / j. issn. 1673-629X. 2020. 08. 034]
　WANG Lian-xi.Sentiment Analysis Method Based on Attribute-sentiment Ontology in Automobile Domain[J].,2020,30(05):193.[doi:10. 3969 / j. issn. 1673-629X. 2020. 08. 034]
[9]黄剑波,陈方灵,丁友东,等.基于情感分析的个性化电影推荐[J].计算机技术与发展,2020,30(09):132.[doi:10. 3969 / j. issn. 1673-629X. 2020. 09. 024]
　HUANG Jian-bo,CHEN Fang-ling,DING You-dong,et al.Personalized Movie Recommendation Based on Sentiment Analysis[J].,2020,30(05):132.[doi:10. 3969 / j. issn. 1673-629X. 2020. 09. 024]
[10]刘昌澍,李　响,詹瑾瑜,等.基于ＴｅｘｔＣＮＮ和ＬｉｇｈｔＧＢＭ 的导游违规行为检测[J].计算机技术与发展,2021,31(05):143.[doi:10. 3969 / j. issn. 1673-629X. 2021. 05. 025]
　,,et al.ＩｌｌｅｇａｌＴｏｕｒＧｕｉｄｅＢｅｈａｖｉｏｒＤｅｔｅｃｔｉｏｎＢａｓｅｄｏｎＴｅｘｔＣＮＮａｎｄＬｉｇｈｔＧＢＭ[J].,2021,31(05):143.[doi:10. 3969 / j. issn. 1673-629X. 2021. 05. 025]

更新日期/Last Update: 2024-05-10