Hits:

Indexed by:会议论文

Date of Publication:2009-09-24

Included Journals:EI、Scopus

Abstract:Research on machine translation has a long history and many methods and techniques have been proposed and developed. However, low quality of translation is still a major problem and many related problems remain unresolved. Super Function based machine translation was proposed to perform translation without going through syntactic and semantic analysis as many machine translation systems usually do. In such way, the import of the errors caused by the syntactic and semantic analysis could be avoided and the readability of the translation is also guaranteed. We continue to extend the Super Function for the Chinese Japanese machine translation. the Super Function is divided into the Super Function for sentences and Super Function for phrases. The existing scope and definition of the variable for Super Function is also expanded. The base noun phrases and base verbal phrases are detected and considered when building the Super Function. The head noun in base noun phrase is treated as variable but the determinative modifiers in base noun phrases are treated as modificatory variables, meanwhile the determinative modifiers in base verb phrases are treated as modificatory variables to build the Super Function. The revised Levenshtein Distance algorithm is proposed for evaluate and find the matching Super Functions in database for input sentences and the determinative modifiers. The proposed extended Super Function keeps the advantages of the Super Function but has less number of Super Functions for Chinese Japanese machine translation. From the experiment results, the extended Super Function is not only effective but also feasible and with considerable precision for Chinese Japanese machine translation. The extended Super Function can even translate long sentences with complex structure for which the completely matching Super Functions can not be found for translation. ?2009 IEEE.