Yes, I do not see an effect... Here is my unit test that test it:
public void testFairSimilarity() throws CorruptIndexException, IOException, ParseException { Directory theDirectory = new RAMDirectory(); Analyzer theAnalyzer = new FrenchAnalyzer(); IndexWriter theIndexWriter = new IndexWriter(theDirectory, theAnalyzer); theIndexWriter.setSimilarity(new FairSimilarity()); Document doc1 = new Document(); Field name1 = new Field(" NAME", "SHORT_SUITE", Field.Store.YES, Field.Index.UN_TOKENIZED); Field content1 = new Field("CONTENT", "x 2 3 4 5 6 7 8 9 10", Field.Store.NO, Field.Index.TOKENIZED); doc1.add(name1); doc1.add(content1); theIndexWriter.addDocument(doc1); Document doc2 = new Document(); Field name2 = new Field(" NAME", "BIG_SUITE", Field.Store.YES, Field.Index.UN_TOKENIZED); Field content2 = new Field("CONTENT", "x x 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20", Field.Store.NO, Field.Index.TOKENIZED); doc1.add(name2); doc1.add(content2); theIndexWriter.addDocument(doc2); Searcher searcher = new IndexSearcher(theDirectory); searcher.setSimilarity(new FairSimilarity()); QueryParser queryParser = new QueryParser("CONTENT", theAnalyzer); Hits hits = searcher.search(queryParser.parse("x")); assertEquals(2, hits.length()); assertEquals("BIG_SUITE", hits.doc(0).get("NAME")); assertEquals("SHORT_SUITE", hits.doc(1).get("NAME")); } Is there anything wrong ? Thanks a lot, Fabrice Daniel Naber-10 wrote: > > On Montag, 21. Januar 2008, Fabrice Robini wrote: > >> I've tried the "fair" similarity described here >> (http://www.nabble.com/a-%22fair%22-similarity-to5806739.html#a5806739) >> with lucene 2.2 but it does not seems to work. > > What exactly doesn't work, don't you see an effect? At least the scores > should change if you try with an artificial small document. Maybe you can > strip down your code to a small self-contained example and post it. > > Regards > Daniel > > -- > http://www.danielnaber.de > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > -- View this message in context: http://www.nabble.com/Is-Fair-Similarity-working-with-lucene-2.2---tp15001250p15014224.html Sent from the Lucene - Java Users mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]