当前位置: X-MOL 学术Ann. Oper. Res. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Identifying Fintech risk through machine learning: analyzing the Q&A text of an online loan investment platform
Annals of Operations Research ( IF 4.4 ) Pub Date : 2020-12-21 , DOI: 10.1007/s10479-020-03842-y
Huosong Xia , Jing Liu , Zuopeng Justin Zhang

Financial risks associated with Fintech have been increasing with its significant growth in recent years. Aiming at addressing the problem of identifying risks in online lending investment under a financial technology platform, we develop a Q&A text risk recognition model based on attention mechanism and Bi-directional Long Short-Term Memory. First, the Q&A pairing on the text data set is carried out, and the matching data set is selected for the next analysis. Secondly, the online loan investment platform is assessed by the named entity recognition of the question text. Finally, the risk level of the corresponding investment platform is evaluated based on the answer text. The experimental results show that the proposed model has achieved improved precision, recall, F1-score, and accuracy compared with other models. Our proposed model can be applied to identify the risks from the text posted on online loan investment platforms and can be used to guide investors’ investment and improve the management of financial technology platforms.



中文翻译:

通过机器学习识别金融科技风险:分析在线贷款投资平台的问答文本

近年来,与金融科技相关的金融风险一直在急剧增长。为了解决金融技术平台下在线贷款投资中识别风险的问题,我们开发了基于注意力机制和双向长期短期记忆的问答文本风险识别模型。首先,对文本数据集进行Q&A配对,然后选择匹配的数据集用于下一个分析。其次,通过问题文本的命名实体识别来评估在线贷款投资平台。最后,根据答案文本评估相应投资平台的风险水平。实验结果表明,与其他模型相比,该模型具有更高的精度,召回率,F1得分和准确性。

更新日期:2020-12-21
down
wechat
bug