Ensemble learning, also known as multiple classifier system, can combine the predictions from multiple base classifiers/learners altogether to conclude a final decision. It has been proven that ensemble learning is a simple, useful and effective meta-classification methodology. SBCB (Selecting Base Classifiers on Bagging) is a selective based ensemble learning algorithm which is able to select an optimal set of classifiers amongst all base classifiers in determining the final result. SBCB equips with an optimization process that is capable of selecting a suitable number of optimal classifiers among all base classifiers automatically, in which, diversity and accuracy are considered as two major selection criteria. In this paper, we built a part-of-speech (POS) tagger for Portuguese based on SBCB learning algorithm as a case study to further investigate the effectiveness and performance of SBCB algorithm. The problem of POS tagging is a practical issue in natural language processing (NLP), especially in the development of a machine translation system. The performance of the POS tagging may interference the subsequent analytical tasks in the translation process, and thereafter affects the translation quality. The POS tagging task can be regarded as a classification problem. Features such as the surrounding context of ambiguous candidates, n-gram information, lexical items and linguistic clues are used and automatically extracted from the Portuguese annotated corpus. The empirical results reveal the effectiveness of SBCB 1 Faculty of Science and Technology, University of Macau Av. Padre Tomás Pereira, Taipa, Macau [email protected], {lidiasc, derekfw }@umac.mo algorithm on POS tagging.