Radu Soricut

I'm a computer scientist working on Machine Translation, Statistical Natural Language Processing, and applications of Machine Learning to natural language tasks. I have done research in areas such as statistical machine translation, language generation, language modeling, syntactic parsing, discourse parsing, question answering, automatic summarization, evaluation of NLP applications, and quality prediction for structured output. I am currently focusing on improving the quality of automatic translations, as well as predicting the quality-level of automatic translations at run-time. For an example of why this is interesting and what the current state-of-affairs is, please visit the Quality Estimation for MT Shared-Task 2012 page. I'm currently a senior research scientist and manager at SDL, in the Language Technologies unit responsible for Statisical Machine Translation (Language Weaver unit). Before that, I was a research scientist at Language Weaver Inc., a Los Angeles--based start-u