Heart Failure is the leading cause of hospitalization among senior adults (older than 65 years old). More than 25% of hospitalized heart failure patients are readmitted within 30 days of discharge, which poses a financial burden and results in a negative quality evaluation for hospitals. Starting on October 1, 2012, the Affordable Care Act (ACA) required the Centre for Medicare and Medicaid Services (CMS) to financially penalize hospitals with an excess of 30-day readmissions.To facilitate efficient and effective 30-day readmission reduction, one of the critical tasks is to identify patients with risk of readmission; We developed one Naïve Bayes risk assessment predictive model and compared it with two commonly used readmission-prediction methods: the HOSPITAL score method and the LACE score method. Our model had a good discriminatory power (AUROC 0.650), and its performance was significantly better than the HOSPITAL score method (AUROC: 0.62, p-value = 0.0003) and LACE score method (AUROC: 0.59, p-value<0.0001). The healthcare utilization data (AUROC: 0.628) achieved the best performance among all the electric health records data.