These complaints are all taken from this dataset of verbal descriptions of medical symptoms and text transcriptions, labeled according to the category of ailment. Each utterance was created by individual human contributors based on a given symptom. The dataset is currently being used to train conversational agents in the medical field.