Proposal for the Establishment of Categories of Linguistic Analysis From the Speech of a Patient with Obsessive Compulsive Disorder, Using Techniques of Automatic Processing of Natural Language

This research aims to set categories of analysis, by exploring the content and linguistic style of an OCD patient, through the computational analysis of transcripts of psychotherapy sessions. Through SketchEngine and LIWC programs three analysis of the patient's speech were applied: a) statistical analysis of cooccurrences of Self, and reanalysis of these results, b) evaluation of functional, content terms by comparison with a dictionary, and c) from an integral perspective we compared the results with three theoretical models of OCD: cognitive-behavioural, systemic and psychodynamic. From the results we propose syntactic and semantic categories based on the observation of words and grammatical constructions, putting emphasis on the review of scientific literature that identifies relevant linguistic patterns of patients with OCD. In short, the establishment of categories based on the analysis of language usage, gives us new arguments to support and reconsider some of the classical postulates attributed to the discourse of these patients. It also reflects on the application of this method to different patients as a tool for better contextualization and patient's evaluation, providing a more personalized and effective objectified measurement methods.