Hindi Analyzer
--------------
Key: LUCENE-2234
URL: https://issues.apache.org/jira/browse/LUCENE-2234
Project: Lucene - Java
Issue Type: New Feature
Components: contrib/analyzers
Reporter: Robert Muir
Priority: Minor
Fix For: 3.1
Attachments: LUCENE-2234.patch
An analyzer for hindi.
below are MAP values on the FIRE 2008 test collection.
QE means expansion with morelikethis, all defaults, on top 5 docs.
||setup||T||T(QE)||TD||TD(QE)||TDN||TDN(QE)||
|words only|0.1646|0.1979|0.2241|0.2513|0.2468|0.2735|
|HindiAnalyzer|0.2875|0.3071|0.3387|*0.3791**|0.3837|0.3810|
|improvement|74.67%|55.18%|51.14%|50.86%|55.47%|39.31%|
* TD was the official measurement, highest score for this collection in FIRE
2008 was 0.3487: http://www.isical.ac.in/~fire/paper/mcnamee-jhu-fire2008.pdf
needs a bit of cleanup and more tests
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-dev-h...@lucene.apache.org
