Translation tool complete

Risk assessment: The model does not show significant improvement in comparison to pure grammar-based models. If the most advanced hybrid model doesn’t fulfill the expectations, MOLTO will have to rely its starting point, purely grammar-based models. These will be sufficient for the case studies in WP6 and WP8, but we may have to cut down the ambitions in WP7. The reasons why our hybrid models failed will be an interesting scientific result anyway, since such models are a focus area in the forefront of machine translation.

Verification: Integrated grammar and STM available.

A one-click system for the hybrid translator has been build and is now available as a shell command on the server in Gothenburg. Partners wishing to test the system should contact UGOT to obtain access to the server.