This report proposes a novel unsupervised document clustering approach based on weighted combination of individual clusterings. Two non-weighted combination methods are adapted to work in a weighted fashion: a graph based method and a probability based one. The performance of the weighted approach is evaluated on real-world collections, and compared to that of individual clustering and non-weighted combination. The results of this evaluation confirm that graph based weighted combination consistently outperforms the other approaches.
CitationGonzález, E., Turmo, J. "Unsupervised document clustering by weighted combination". 2006.
All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder. If you wish to make any use of the work not provided for in the law, please contact: email@example.com