Machine learning and natural language processing in psychotherapy research: Alliance as example use case.

摘要: Artificial intelligence generally and machine learning specifically have become deeply woven into the lives technologies of modern life. Machine is dramatically changing scientific research industry may also hold promise for addressing limitations encountered in mental health care psychotherapy. The current paper introduces natural language processing as related methodologies that prove valuable automating assessment meaningful aspects treatment. Prediction therapeutic alliance from session recordings used a case point. Recordings 1,235 sessions 386 clients seen by 40 therapists at university counseling center were processed using automatic speech recognition software. algorithms learned associations between client ratings exclusively linguistic content. Using portion data to train model, modestly predicted content an independent test set (Spearman's ρ = .15, p < .001). These results highlight potential harness predict key psychotherapy process variable relatively distal Six practical suggestions conducting are presented along with several directions future research. Questions dissemination implementation be particularly important explore improves its ability automate outcome. (PsycInfo Database Record (c) 2020 APA, all rights reserved).