A Credit Evaluation Model For Small Micro-enterprises Based On Text Mining And Empirical Study

According to credit reports of small micro-enterprise in the bank, this paper verifies the authenticity of credit indexes based on text mining. And credit indexes are selected with PCA. In order to compare, credit evaluation models are built respectively with indexes corrected and non-corrected. The result shows that credit evaluation model built on authenticity corrected with text mining could use fewer indexes to evaluate small micro-enterprise credit effectively. And the evaluation model could guide the bank to establish different competition strategies of loan price. There are more qualitative indexes than financial indexes in small micro-enterprise credit evaluation model. It suggests that more attention should be paid to development prospect of the small micro-enterprise when its credit is evaluated.