Emmanuel Castro Santana wrote:
Hi you all
We are trying to set up the schema.xml for a better similarity search using
MoreLikeThisHandler and we wonder what this termOffsets option is actually
used for. All we know so far is that it increases the storage costs, but how
exactly ?
We have been struggling to find this information for a while, if any of you
could give us a help I would really appreciate.
Thanks in advance.
Emmanuel Santana
Briefly looking at MLT source code, I don't think it uses termOffsets,
but it uses termVectors for better performance when MLT constructs
a BooleanQuery which is used to search similar documents.
Koji
