Security bug report (SBR) prediction has been increasingly investigated for eliminating security attack risks of software products. However, there is still much room for improving the performance of automatic SBR prediction. This work is inspired by the work of two recent studies proposed by Peters et al. and Wu et al., which are focused on SBR prediction and both published on the top tier journal TSE (Transactions on Software Engineering). The goal of this work is to improve the effectiveness of supervised machine learning-based SBR prediction with the help of software security domain knowledge. It first extracts software security domain knowledge from CWE (Common Weakness Enumeration) and CVE (Common Vulnerabilities Exposure), which are authoritative sources of software vulnerability. After that, the matrix of bug reports is generated based on the roots of security domain keywords. Large-scale experiments are conducted on a set of trustworthy datasets cleaned by Wu et al. The results show our domain knowledge-guided approach could improve the effectiveness of SBR prediction by 25% in terms of F1-score on average.