I have made the code to find the similarity between two documents by finding their tf and then their cosine values . But when i was looking at the standard examples on lucene , every program had made use of an index . My process involves a comparision between one reference document and other documents from a folder .
Do u think i should use indexing ?