PsyCredit: An interpretable deep learning-based credit assessment approach facilitated by psychometric natural language processing
作者:
Highlights:
• A novel deep learning-based framework with social media data, i.e., PsyCredit, is proposed for credit risk assessment.
• A novel deep learning-based psychometirc approach is proposed to identify the personality traits of individuals through their social media postings.
• An interpretable module with the deep learning technique can interpret the underling mechanisms.
• Our proposed framework verifies the predictive power of personality traits from social media on individuals' credit risk.
摘要
•A novel deep learning-based framework with social media data, i.e., PsyCredit, is proposed for credit risk assessment.•A novel deep learning-based psychometirc approach is proposed to identify the personality traits of individuals through their social media postings.•An interpretable module with the deep learning technique can interpret the underling mechanisms.•Our proposed framework verifies the predictive power of personality traits from social media on individuals' credit risk.
论文关键词:Credit risk,Personality traits,Social media,Psycholinguistic,Deep learning,Explainable artificial intelligence
论文评审过程:Received 7 February 2021, Revised 22 November 2021, Accepted 6 March 2022, Available online 12 March 2022, Version of Record 15 March 2022.
论文官网地址:https://doi.org/10.1016/j.eswa.2022.116847