Abstract
Due to the technological progress and the continuous upload on the Web, an enormous amount of documents has been accumulating. This accumulation became an issue since it makes the data big and its mining difficult. Therefore, the focus of this work is the extraction of useful data in terms of quality and time by ameliorating search results. In this paper, we propose a framework that eliminates the duplications in the first place, then making use of a clustering algorithm combined with a distance measure filters and classifies the results in order to reduce the amount of documents efficiently and gain in terms of documents quality and search time. The proposed architecture is based on k-means clustering algorithm and the cosine similarity measure. The system showed encouraging results.