This study proposes a novel approach to attribute clustering. It exploits the strength of semi-supervised learning to improve the quality of attribute clustering particularly when labeled data is limited. The significance of this work derives in part from the broad, and increasingly important, usage of attribute clustering to address outstanding problems within the machine learning community. This form of clustering has also been shown to have strong practical applications, being usable in heavyweight industrial applications. Although researchers have focused on supervised and unsupervised attribute clustering in recent years, semi-supervised attribute clustering has not received substantial attention. In this research, I propose an innovative two step iterative semi-supervised attribute clustering framework. This new framework, in each iteration, uses the result of attribute clustering to improve a classifier. It then uses the classifier to augment the training data used by attribute clustering in next iteration. This iterative framework outputs a better classifier and a better attribute clustering at the same time. Compared with the few existing semi-supervised approaches, it gives more accurate clusters of attributes which better fit the real relations between attributes. In experiments, I demonstrated the power of this approach in attribute selection and automatic definition of subspaces for a subspace-based ensemble. Every usage of attribute clustering could be used to set up this framework. As future work I will test this framework in automatic definition of views in co-training and in missing attribute value handling.