JIRA issues opened for the project
Issue |
Description |
Status |
[GSoC] Implementing State of the Art Ranking for Lucene |
All sub-tasks completed |
|
Rename some classes and methods in the search package so their names describe their roles |
Tentative |
|
Make a Stats class |
Done |
|
Implement various ranking models as Similarities |
Done |
|
Unit and integration test cases for the new Similarities |
Done |
|
EasySimilarity to interpret document length as float |
Done |
|
Integrate MockBM25Similarity and MockLMSimilarity into the framework |
Done |
|
Rename EasySimilarity to SimilarityBase |
Done |
|
Make EasySimilarityProvider a full-fledged class |
Done |
|
Get javadoc for the similarities package in shape |
Done |
Ideas
|
Introduce NormConverter so that any Similarity can be used on any index? |
Under consideration |
General observations
Issue |
Description |
Status |
|
The javadoc for IndexReader.totalTermFreq(), and its callee is just a copy of the description of IndexReader.docFreq(), even though the semantics of the method differ |
Not opened yet |