Abstract:
In machine translation, the resolution of all types of anaphora is not needed. It is classified as translation anaphora and non-translation anaphora. The main steps of the text-based machine translation are the identification of discourse units, search for possible anaphoric devices (ADs), search for the corresponding antecedent to each AD, replacement of each AD by its actual antecedent, dissection of each compound and complex discourse into primitive discourses, conversion of text into source language representation (SLR), transfer of SLR into target language representation (TLR), conversion of TLR into primitive discourses and finally pronominalisation of the target text. These steps are the focal point of this paper.
Page(s):
121-134
DOI:
DOI not available
Published:
Journal: Scientific khyber, Volume: 7, Issue: 2, Year: 1994