Docket #: S15-036
Phrasal: A Statistical Phrase-based Machine Translation System
FOR ALL LICENSING INQUIRIES PLEASE EMAIL NLP Licensing.
Phrasal is a statistical phrase-based machine translation system. It provides all components necessary to build a state-of-the-art translation system from raw text. Recent improvements include support for training large-scale, feature-rich models, and for interactive machine translation. Unlike Moses and cdec, two popular MT toolkits, Phrasal is written in Java, so it can be deployed in J2EE production environments.
This technology is available under GPL here
If a commercial license under alternative terms is desired then please contact the licensing contact below.
Applications
- Automatic human language translation.
Advantages
- Fast, online tuning of feature-rich translation models
- Interactive decoding for interactive machine translation
Similar Technologies
-
Fast feature selection method and system for maximum entropy modeling S04-319Fast feature selection method and system for maximum entropy modeling
-
Methods for Measuring Diversity using Multi-Dimensional Networks S20-484Methods for Measuring Diversity using Multi-Dimensional Networks
-
State-of-the-Art Graph Diffusion Transformer for Natural Language Processing S20-271State-of-the-Art Graph Diffusion Transformer for Natural Language Processing