With the growing popularity of flea market applications, appropriate pricing becomes a critical factor in facilitating smooth transactions. In particular, business books are a highly demanded category, but their pricing is influenced by various factors such as book condition, publication year, popularity, and author recognition. This study aims to develop a predictive model for estimating the price ratio (selling price/original price) of business books based on actual C2C transaction data. To achieve this, we construct a regression model using Random Forest and Logistic Regression, incorporating multiple features related to book attributes and transaction details. Furthermore, SHAP analysis is employed to interpret feature contributions and identify key factors influencing price formation. The results demonstrate the feasibility of price prediction and provide insights that can assist users in setting reasonable prices.