Multi-label classification is the supervised learning problem where an instance may be associated with multiple labels. This is opposed to the traditional task of single- label classification (i.e. multi-class, or binary) where each instance is only associated with a single class label. The multi-label context is receiving increased attention and is applicable to a wide variety of domains, including text classification, scene and video classification, and bioinformatics. The Binary Relevance (BR) used the commonly used the multi-label classification. This approach address this problem by indicating a separate classifier for each class. The existing techniques makes prone to error. The proposed system used the two technique used to reduce the unnecessary dependencies and to reduce error-propagation. That new classifier-stacking technique name PruDent. Nowadays, multi-label classification methods are increasingly required by modern applications, such as protein function classification, music categorization and semantic scene classification. The straightforward approach to multi-label classification is based on decomposition, which essentially treats all labels independently and ignores interactions between labels. Many approaches to MLC take a decompositive approach, i.e., they decompose the MLC problem into a series of ordinary classification problems. The technique PruDent, proposed in this paper, seeks to address both of these issues. Unnecessary label dependencies are pruned out, and confidence values calculated for the classifications are employed in a manner that reduces error propagation. In this project we propose a new adaptation for the Binary Relevance method taking into account the correlation among labels, focusing on the interpretability of the model, not only its performance.