lucene-使用sandbox的wordnet完成同义词索引

wordnet是一个同义词系统。

1、从wordnet网站下载解压prolog16.tar.gz文件

2、获得sandbox工具箱wordnet包的二进制文件

3、建立同义词索引

java org.apache.lucene.wordnet.Syns2Index prologwn/wn_s.plwordnetindex

4、将WordNet中的同义词连接到分析器中

public class WordNetSynonymEngine implements SynonymEngine{

    RAMDirectory directory;

    IndexSearcher searcher;

    

    public WordNetSynonymEngine(File index) throws IOException{

            directory=newRAMDirectory(FSDirectory.getDirectory(index,false));

            searcher=new IndexSearcher(directory);

    }

    public String[] getSynonms(String word) throws IOException{

         ArrayList synList=new ArrayList();

         Hits hits=searcher.search(new TermQUery(newTerm("word",word)));

          

         for(int i=0;i<hits.length();i++){

            Document doc=hits.doc(i);

            String[] values=doc.getValues("syn");

            for (int j=0;j<values.length;j++){

                 synList.add(values[j]);

             }

         }

        return (String[]) synList.toArray(new String[0]);

    }

}