Select a language, enter text, and run all 4 models (MED, DIS, SYMP, PROC).
Lower = more entities predicted