Identifying the implementation of neural network approaches in peer-to-peer lending research: a bibliometric-based thematic approach

Peer-to-peer (P2P) lending market has exploded in popularity since the last decade. The proliferation of data has given opportunities to prediction models, such as neural network (NN), to analyse and forecast risk assessment. The objective of this research is to explore the intersection of NN models in P2P lending and identify future trends for NN in this field. A systematic literature review (SLR) was conducted using the PRISMA model and bibliometric analysis, which included network and thematic investigation approaches for the NN in P2P lending research published over the last decade. The study analysed the key trends in select research domains, identifying four themes: predictive analysis, financial risk, convolutional neural networks, and P2P networks. The research also identified citation networks with four clusters: investor behaviour, borrower behaviour, classification models for credit scoring, and borrower default prediction. Further, analysis was performed on the most cited documents, emphasising the research methods, models, and datasets used in the articles.