lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Klaus" <>
Subject Similarity
Date Mon, 19 Dec 2005 18:23:12 GMT
Hi All,


I'm new to lucene and a have some questions according to the entire system.


I) What is exactly written to the index? Is the index just an inverted list?
Is there term weight scoring stored?


II) How works the retrieval process work? I guess so:


1)       Get all the documents from the index via the inverted list. 

2)       Compute the score for every document and the query with the
similarity class. As far as i can see, the similarity is just based on the
tf-ddf weighting? Is there no cosine measure or so used, to compare the
document and the query vector?


Thanks a lot



  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message