Identifying Technical Debt (TD) within Software Development Projects (SDP) is a growing research interest as it has the potential to save software developers’ effort in maintenance tasks. Although there are ten types of TD, there is a lack of automatic techniques to extract them through static nor dynamic analysis. This paper proposes a self-admitted TD extraction framework to extract TD from software’s comments and classify them comprehensively through a Parts-of-Speech technique. A public TD dataset is used to evaluate the proposed framework. Results show that the proposed technique was able to increase the classification of build, architectural, versioning, and infrastructure TD by 16%.