Recent years are considered to be the ‘golden age’ of data science and machine learning. Impressive improvements in both software and hardware capabilities, along the last decade, allow researchers to handle broad set of problems which were unsolved until lately. The explosion of available data, in various forms, is a central aspect in such revolution due to the basic concept that larger dataset enables to train and validate stronger machine learning models. Central branch in machine learning is natural language processing (NLP), dealing with various problems where algorithm’s input is in the form of textual data.

A lot the NLP research is done for specific domains (e.g., social media, low & legal). Mental health diagnosis and care is one of them, having its specific algorithms developed, aiming to solve central tasks in the domain (e.g., detect symptoms of severe mental illness).

In this talk I will provide a brief introduction to relevant machine learning terms and algorithm types. The rest of the talk will be dedicated to map NLP relevant tasks and algorithms (e.g., classification, translation) and explain how such can be relevant and be used by the mental health and care community. Along mapping such tasks and algorithms, I will point out to the latest papers published in the field of NLP for mental health diagnosis and care.