Accurate information retrieval using text analysis and disambiguation

Abstract

The most common method for retrieving information from text is the string search. In this method, search is targeted to the string in text. This string can consist of a single word, word clusters, partial words etc. Search results are often faulty. The result may have such hits that were not intended. Also, part of targeted words was not retrieved. These problems can be largely overcome by including language analysis into the search system. The report discusses two types of information retrieval implementation, where language analysis and disambiguation are included.