Y. Nitta (Japan)
Advanced Technology in Education and Training, Computer-Assisted Learning and Instruction, Interactive Learning Environments, Educational Software, Machine Translation, E-Learning
We investigate the method to reinforce the performance of language e-learning using the findings and linguistic knowledge of machine translation study. Especially the classical transfer-base machine translation systems can provide powerful knowledge resources for language e learning. The classical transfer-base machine translation has sound and transparent mechanism, which is mainly coming from the carefully designed linguistic patterns. In our CREST MT research, we have newly designed linguistic patterns, which can represent not only the surface sentential or syntactic patterns but also the semantic patterns such as predicate augment structure. The pattern transformation is carried out through the concept of “semantically equivalent transformation”, which is a mechanism for preserving logical semantics of each sentence. We will show some examples of applying patterns to language e-learning.
Important Links:
Go Back
