TWIRL: Translation with rule learning

Machine translation (MT) is the automatic conversion of a source language to a target language using computers. The two most common paradigms for machine translation are the rule based and example based approaches. The problem with the example based approach is that it needs to be domain specific and a large database of examples is needed to produce accurate translation results. Rule based approaches are known to produce high quality translations however, a linguist is necessary in deriving the set of rules. To overcome the problems of both the example based and the pure rule based paradigms, TWiRL used the rule based approach with an integration of machine learning of rules to allow flexibility in translation. Since no rule learning has been explored in English to Filipino machine translation system, the focus of this research is on translating English to Filipino text.Keywords: Rule generalization, Compositionality, Bilingual Corpus, rule based, Machine Translation, Seed Rule Generation.