I try to use Lucene to make some experiments with the distribution of words
in documents. Using the TermPositionVector, one can compute some statistics
about word positions (mean, standard deviation, etc.).
It is possible to store such statistical values directly in the Lucene
For example, I want to save the statistical values of each Term by the index
process, and get them later in query time.
Term => <docNum, freq, <X1, X2, ... ,Xn> >
Where <X1, X2, ... ,Xn> are some statistical values about the Term position
Should I modify the classes: IndexWriter.java and IndexReader.java?