Agglomerative Clustering

This book presents an agglomerative clustering method modelling the data as a graph. An evolutionary algorithm is employed on the graph for finding the components which represent a first level clustering. Then, an agglomerative procedure is used for refining the first solution, achieving a noticeable improvement in the final solution. A case study is proposed for testing the method, which is the discrimination of text documents according to different languages. An experiment shows that this method is quite promising in discriminating among the documents in the different languages.