Title: ANN and case-based reasoning methods for English to Arabic machine translation

Authors: Marwan Akeel; R.B. Mishra

Addresses: Department of Computer Engineering, Indian Institute of Technology-Banaras Hindu University (IIT-BHU), Varanasi-221005 U.P., India ' Department of Computer Engineering, Indian Institute of Technology-Banaras Hindu University (IIT-BHU), Varanasi-221005 U.P., India

Abstract: The basic assumption of using case base reasoning in machine translation is that similar sentences in source language have similar translation in the target language. In our approach, English input sentences with similar syntactical features have similar syntactical features at its Arabic output translation. The syntactical features of the new English input sentence is matched with old cases from the stored case base using artificial neural network method. We use feed forward artificial neural network for the selection of the Arabic syntactical features when the input is an English sentence structure and attributes. The retrieved case is adapted using rules.

Keywords: case-based reasoning; CBR; machine translation; artificial neural networks; ANNs; rule-based approach; English to Arabic translation; translation rules; syntax.

DOI: 10.1504/IJKESDP.2014.069294

International Journal of Knowledge Engineering and Soft Data Paradigms, 2014 Vol.4 No.4, pp.371 - 397

Published online: 09 May 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article