Fig. 4From: CERC: an interactive content extraction, recognition, and construction tool for clinical and biomedical textThe automatic language translator design schema. The system utilizes a machine learning environment that incorporates CERC, Google Translate, a “self-learning translator,” and “a language library”. CERC first processes narrative text from de-identified discharge summaries, Google Translate then translates the resulting summary from English into different languages. Professional language experts correct the translated text and the self-learning translator takes in the processed discharge text as well as the expert corrected content, learns adaptively from the corrections and retains that knowledge in its self-learning libraryBack to article page