Tue, 08 Jan 2013 07:08:00 GMT

public class ClusteringFlyStoneDocument {

    public static void main(String[] args) {

        final Controller controller = ControllerFactory.createCachingPooling(IDocumentSource.class);
        //创徏需要处理的属性对�?/span>
        final Map attributes = Maps.newHashMap();

        //需要处理的文档集合
        List documents = SampleDocumentData.DOCUMENTS_DATA_MINING;

        //��文档添加到聚类属性中
        CommonAttributesDescriptor.attributeBuilder(attributes).documents(documents);

        //讄��处理的语�a��Q�）
        MultilingualClusteringDescriptor.attributeBuilder(attributes).defaultLanguage(LanguageCode.CHINESE_SIMPLIFIED);

        //讄��需要处理的对象�Q�以及聚�cȝ��法
        final ProcessingResult englishResult = controller.process(attributes, LingoClusteringAlgorithm.class);

        //获取�l�果�Q�打印聚�c�M��题及兌��信息
        final List clustersByTopic = englishResult.getClusters();
        System.out.println("=======聚类主题=====");
        for(Cluster cluster : clustersByTopic){
            System.out.println("【主�?nbsp;�? +cluster.getLabel());
            List cDocLst = cluster.getAllDocuments();
            for(Document doc : cDocLst){
                System.out.println("--------" + doc.getTitle());
            }
        }

        //通过URL�q�行聚类
        final ProcessingResult byDomainClusters = controller.process(documents, null,ByUrlClusteringAlgorithm.class);
        final List clustersByDomain = byDomainClusters.getClusters();
        System.out.println("=======URL聚类=======");
        for(Cluster cluster : clustersByDomain){
            System.out.println("【URL�? +cluster.getLabel());
            List cDocLst = cluster.getAllDocuments();
            for(Document doc : cDocLst){
                System.out.println("----" + doc.getTitle());
            }
        }
    }

}

沙漠中的�?/a> 2013-01-08 15:08 发表评论

思热99re视热频这里只精品,亚洲精品在线三区,国产98在线