A Survey of Document Search Using Images

Abstract

Web search engine is software system that is designed to search for information on WWW. When user gives
a query into search engine, it provides to the user, a list of best matching web pages according to its criteria, generally
with a short summary consisting the documents title and part of the text. A picture convey information more efficiently
and effectively comparative to words. Traditional text based search engine is based on textual similarity. Web search
engines do not use the pictures in the web pages to find relevant documents for a given user query. Instead of images,
they operate by computing a measure of consistency between only the text portion of each page and the keywords
provided by the user. This paper presents survey of various techniques which are used for re-ranking. It introduces the
information about existing approaches towards the re-ranking techniques