Translation Pattern Extraction and Recombination for Example-Based Machine Translation

An approach to Example-Based Machine Translation is presented which operates by extracting and recombining translation patterns from a bilingual corpus aligned at the level of the sentence. The translation patterns are extracted using a recursive machinelearning algorithm based on the principle of similar distributions of strings: source and target language lexical items that co-occur in the same two sentence-pairs are likely to be translations of each other. The translation patterns extracted represent generalisations of sentences that are translations of each other in that certain sequences of words are replaced by variables. The translation patterns resemble, to a certain extent, transfer rules but with less constraints since there is no concept of syntactic structure in this approach: translation patterns are extracted based on the information inherent in the corpus. The strings and variables of which the translation patterns are composed are aligned in order to provide a more refined bilingual knowledge source, necessary for the recombination phase where target language translations are produced. They are aligned by means of language-neutral techniques – cognates and bilingual lexical distribution – in order to maintain language-neutrality and portability. A non-structural approach based on the distributions of surface forms in a corpus is error prone and liable to the extraction of translation patterns that are false translations. This thesis highlights some of the sources of those errors and proposes solutions to them based on the addition of external linguistic knowledge. First, the basic approach is augmented with morphological analysis. Second, part-of-speech tagging is incorporated. This results in three variants of the approach, each with increasing amounts of external linguistic resources included. The performance of each variant is assessed and a comparison is made between them in order to test the hypothesis that an increase in linguistic knowledge improves performance, in terms of both recall and precision.