I am doing clustering on twitter data set using k-means. when I calculate the silhouette for the clustering, I got very low score. that is meaning the clustering is not good