Evaluation of Text Clustering Algorithms with N-Gram-Based Document Fingerprints