Natural language processing (NLP) techniques can be used to make inferences about peoples' mental states from what they write on Facebook, Twitter and other social media. These inferences can then be used to create online pathways to direct people to health information and assistance and also to generate personalized interventions. Regrettably, the computational methods used to collect, process and utilize online writing data, as well as the evaluations of these techniques, are still dispersed in the literature. This paper provides a taxonomy of data sources and techniques that have been used for mental health support and intervention. Specifically, we review how social media and other data sources have been used to detect emotions and identify people who may be in need of psychological assistance; the computational techniques used in labeling and diagnosis; and finally, we discuss ways to generate and personalize mental health interventions. The overarching aim of this scoping review is to highlight areas of research where NLP has been applied in the mental health literature and to help develop a common language that draws together the fields of mental health, human-computer interaction and NLP.

No abstract

This paper was originally published in Biological Procedures Online (BPO) on March 23, 2006. It was brought to the attention of the journal and authors that reference 74 was incorrect. The original citation for reference 74, "Stanford V. Biosignals offer potential for direct interfaces and health monitoring.

Abstract. It is widely acknowledged that learners experience a variety of emotions while interacting with Intelligent Tutoring Systems (ITS), hence, detecting and responding to emotions might improve learning outcomes. This study uses machine learning techniques to detect learners' affective states from multichannel physiological signals (heart activity, respiration, facial muscle activity, and skin conductivity) during tutorial interactions with AutoTutor, an ITS with conversational dialogues. Learners were asked to self-report (both discrete emotions and degrees of valence/arousal) the affective states they experienced during their sessions with AutoTutor via a retrospective judgment protocol immediately after the tutorial sessions. In addition to mapping the discrete learning-centered emotions (e.g., confusion, frustration, etc) on a dimensional valence/arousal space, we developed and validated an automatic affect classifier using physiological signals. Results indicate that the classifier was moderately successful at detecting naturally occurring emotions during the AutoTutor sessions.