Identification of syntactic ambiguities in Pashto text

Abstract

Natural languages are inherently ambiguous. For a machine translation system, it is an essential task to resolve the ambiguities in the source language. Before resolution of ambiguities, their identification is an essential task. This research paper is about the classification and identification of syntactic ambiguities in Pashto text. Here, the identification is based on the parse trees build by the parser for the ambiguous phrases. An algorithm is proposed which correctly identifies syntactic ambiguities with a success rate of 64.3%.