In recent years, with the development of science and technology, there were considerable advances in datasets in various sciences, and many features are also shown for these datasets nowadays. With a high-dimensional dataset, many features are generally redundant and/or irrelevant for a provided learning task, which has adverse effects with regard to computational cost and/or performance. The goal of feature selection over partially labeled data (semi-supervised feature selection) is to choose a subset of available features with the lowest redundancy with each other and the highest relevancy to the target class, which is the same objective as the feature selection over entirely labeled data. By appropriate reduction of the dimensions, in addition to time-cost savings, performance increases as well. In this paper, side information such as pairwise constraint is used to rank and reduce the dimensions. In the proposed method, the authors deal with checking the quality (strength or uncertainty) of the pairwise constraint. Usually, the quality of the pair of constraints on the dimension reduction is not calculated. In the first step, the strength matrix is created through a similarity matrix and uncertainty region. And then, by using the strength and similarity matrices, a new constraint feature selection ranking is proposed. The performance of the presented method was compared to the performance of the state-of-the-art, and well-known semi-supervised feature selection approaches on eight datasets. The findings indicate that the proposed approach improves previous related approaches with respect to the accuracy of constrained clustering. In particular, the numerical results showed that the presented approach improved the classification accuracy by about 3% and reduced the number of selected features by 1%. Consequently, it can be said that the proposed method has reduced the computational complexity of the machine learning algorithm despite increasing the classification accuracy.