Kernel Regression Based Machine Translation

We present a novel machine translation framework based on kernel regression techniques. In our model, the translation task is viewed as a string-to-string mapping, for which a regression type learning is employed with both the source and the target sentences embedded into their kernel induced feature spaces. We report the experiments on a French-English translation task showing encouraging results.

• Type: Conference
• Year: 2007
• Where: NAACL
• Authors: Zhuoran Wang, John Shawe-Taylor, Sándor Szedmák
• Keywords: Computational Linguistics, Kernel Regression Techniques, Machine Translation Framework, NAACL 2007, Translation Task